Macrophage-Membrane-Coated Nanoparticles for Tumor-TargetedChemotherapyYu Zhang,† Kaimin Cai,‡ Chao Li,† Qin Guo,† Qinjun Chen,† Xi He,† Lisha Liu,† Yujie Zhang,†

Yifei Lu,† Xinli Chen,† Tao Sun,† Yongzhuo Huang,§ Jianjun Cheng,*,‡ and Chen Jiang*,†

†Key Laboratory of Smart Drug Delivery, Ministry of Education, State Key Laboratory of Medical Neurobiology, Department ofPharmaceutics, School of Pharmacy, Fudan University, Shanghai 201203, China‡Department of Materials Science and Engineering, University of Illinois at Urbana−Champaign, Urbana, Illinois 61801, UnitedStates§Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 501 Haike Road, Shanghai 201203, China

*S Supporting Information

ABSTRACT: Various delivery vectors have been integrated within bio-logically derived membrane systems to extend their residential time andreduce their reticuloendothelial system (RES) clearance during systemiccirculation. However, rational design is still needed to further improve the insitu penetration efficiency of chemo-drug-loaded membrane delivery-systemformulations and their release profiles at the tumor site. Here, a macrophage-membrane-coated nanoparticle is developed for tumor-targeted chemotherapydelivery with a controlled release profile in response to tumor microenviron-ment stimuli. Upon fulfilling its mission of tumor homing and RES evasion,the macrophage-membrane coating can be shed via morphological changesdriven by extracellular microenvironment stimuli. The nanoparticles dis-charged from the outer membrane coating show penetration efficiencyenhanced by their size advantage and surface modifications. After internal-ization by the tumor cells, the loaded drug is quickly released from thenanoparticles in response to the endosome pH. The designed macrophage-membrane-coated nanoparticle (cskc-PPiP/PTX@Ma) exhibits an enhanced therapeutic effect inherited from both membrane-derived tumor homing and step-by-step controlleddrug release. Thus, the combination of a biomimetic cell membrane and a cascade-responsive polymeric nanoparticle embodiesan effective drug delivery system tailored to the tumor microenvironment.

KEYWORDS: tumor microenvironment, macrophage-membrane coating, cascade-responsiveness, biomimetic delivery system,breast-cancer targeting

Nanoparticles have been explored as a promising deliveryvector for cancer therapeutic agents with the potential for

great impact on future public health.1,2 To achieve favorableantineoplastic effects, the formulations should cross multiplephysiological barriers by responding appropriately to differentintracorporal milieus. A nanoparticle requires long residence insystemic circulation with considerable stability against plasmadilution, opsonization, and reticuloendothelial system (RES)clearance.3,4 Upon passive or active accumulation at acancerous site, the tumor uptake amount and drug-releaseefficiency will be the key factors in its general curative effect.Therefore, the rational design of drug delivery systems mustincorporate both stabilizing strategies and on-demand drug-release mechanisms. Miscellaneous approaches involvingparticle size, surface charge, morphology, and terminalmodification have been investigated to extend circulation half-life and improve targeting ability. Nevertheless, syntheticnanomaterials have still been reported to possess dangerous

characteristics or behaviors such as immunological responseand off-target protein adsorption during systemic circulation.5,6

Nanoparticles camouflaged by biologically derived cellmembranes have recently attracted attention for theirprolonged circulation in vivo.7,8 Their innate self-recognitionfeatures optimize their dynamic properties by cloaking thecontents to evade RES elimination and immunologicalsurveillance. The membrane camouflage strategy can achievea superior half-life in systemic circulation and improve tumoradhesion ability without loss of drug-loading capacity or thenanosize advantage.9,10 Various delivery vectors including gold-based nanoplatforms, upconversion nanoparticles, and meso-porous silica nanoparticles have been integrated into macro-phage-membrane systems with good stability during systemiccirculation.11,12 The innate inflammation-directed chemotactic

Received: December 14, 2017Revised: February 21, 2018Published: February 23, 2018

Letter

pubs.acs.org/NanoLettCite This: Nano Lett. 2018, 18, 1908−1915

© 2018 American Chemical Society 1908 DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

Dow

nloa

ded

via

UN

IV I

LL

INO

IS U

RB

AN

A-C

HA

MPA

IGN

on

Oct

ober

26,

201

8 at

19:

56:1

0 (U

TC

).

See

http

s://p

ubs.

acs.

org/

shar

ingg

uide

lines

for

opt

ions

on

how

to le

gitim

atel

y sh

are

publ

ishe

d ar

ticle

s.

ability of macrophages could drive the vector to accumulate inchronic inflammatory tumor tissue, offering promise fororthotopic therapies such as photothermic and photodynamictherapy.13,14 However, barriers to chemo-drug delivery bymacrophage-membrane-camouflaged systems remain. In theclassical mononuclear phagocytic system, macrophages areinherently driven to eliminate malignant cells and depletefibrosis.15,16 However, contrary to its phagocytotic nature, themacrophage-derived vesicle must instead be engulfed by tumorcells to achieve drug delivery. After accumulation at tumor sites,the membrane coating becomes an obstacle impeding drugrelease. Therefore, an appropriate membrane escape tactic is anadditional prerequisite for chemo-drug delivery platforms withbiologically derived membrane coatings.Meanwhile, we noted that a similar membrane escape

phenomenon had been discovered in the intracellular releaseprocess in endosomes. Cationic polymeric drug carriers, such asH+-capturing sponges, can induce an excessive influx ofelectrolyte and water into the acidic late endosome to achievean internal equilibrium of electric neutrality and ion strength,thus impelling the endosome membrane to burst and causingextravasation of the nanoparticle contents.17,18 This phenom-enon is termed the proton sponge effect.19 Our previous studieson poly(amino acid)-based vectors also showed favorableendosome escape and intracellular drug-release abilities.20−22

Like the endosomal environment, the tumor microenvironmentfeatures reduced pH owing to oncogenic transformation andabnormal metabolism.23,24 Thus, the acidic tumor-tissuemicroenvironment may provide H+-rich conditions enablingnanoparticle escape from the membrane enclosure, similar tothat occurring in the late endosome via the proton spongeeffect.Herein, inspired by the proton sponge effect, a biomimetic

macrophage-membrane-coated nanoparticle (cskc-PPiP/PTX@Ma) was engineered to exhibit step-by-step release behavior inresponse to the differences in pH in the tumor microenviron-ment. Natural macrophage membranes with their associatedmembrane proteins were reconstructed into vesicles withoutloss of their inflammatory tumor-homing ability. Duringsystemic circulation, the vesicle membrane was expected toserve as a concealing cloak against opsonization and RESclearance and as a tumor-homing navigator to enhance tumoraccumulation. In the first release stage, once the tumor-targeting task of the membrane cloak was complete, theinterstitial pH would cause the membrane-coated formulationto undergo expansion and eruption, removing the coat. Thedischarged nanoparticles could then be further taken up bytumor cells, assisted by their surface modification with atargeting peptide. In the second release stage, the encapsulatedchemo-drug would finally be released from the nanoparticles inresponse to the intracellular pH of the tumor cells. This step-by-step release strategy should optimize the drug-releasekinetics in the tumor microenvironment while maintainingthe tumor-targeting ability of the membrane-coating system insystemic circulation. Paclitaxel (PTX), a classic hydrophobicanticancer drug, was chosen as the model drug to test theformulation’s delivery capacity and therapeutic effect in anorthotopic breast-cancer-bearing mouse model. The biodis-tribution, pharmacodynamics, and general tissue toxicitysuggested that the resulting formulation, cskc-PPiP/PTX@Ma, offered promise for breast-cancer therapy.The preparation of the membrane-coated nanoparticles is

illustrated in Figure 1A. For the construction of the internal

nanoparticle, amphiphilic bola-pattern polymers with selectedside chains were synthesized by the dual-end PEGylation ofpoly(β-amino ester) for hydrophobic drug loading.25,26 ThepH-sensitive polymer was functionalized with a cationic 2-aminoethyldiisopropyl group (PPiP) to tune its buffer capacityto the extracellular pH of the tumor region. The pH-insensitivepolymer with neutral octyl group side chains (PPC8) sharedthe same carbon number but lacked the branched structure andprotonation capacity of the pH-sensitive polymer (Scheme S1,Table S1). The molar ratio of backbone to side chain monomerwas tuned to 1.05:1 to produce a symmetric structure with bothends terminating in acrylate groups for further terminalPEGylation. To further facilitate tumor cell uptake aftertumor homing and membrane exuviation, a synthetic D-formoligopeptide with the sequence cskc was chosen as the targetingligand for nanoparticle surface modification. This oligopeptidewas reported to show high affinity to the insulin-like growth

Figure 1. A. Scheme of the preparation of membrane-coatednanoparticles. B. Microscope images (63× oil lens, crop) of cskc-PPiP/PTX@Ma (a) and PPC8/PTX@Ma (b) in buffers at pH 7.4 and6.5.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1909

factor 1 receptor (IGF1R), which is aberrantly highly expressedon tumor cells because of metabolic disturbance.27−29 Thetargeting polymer (cskc-PPiP) was optimized by terminalconjugation of the IGF1R-targeting peptide on the hydrophilicPEG end of PPiP. All the polymers were characterized by 1HNMR spectroscopy (Figures S1−S6). The membrane-coatednanoparticles were prepared by repeated extrusion of the as-

prepared nanoparticle with freshly extracted macrophagemembrane and were proven to possess sufficient drug-loadingcapacity for in vivo drug delivery (Table S2). The morphologyof cskc-PPiP/PTX@Ma consisted of uniform spheres at pH 7.4but cracked at pH 6.5 to give sickle-shaped particles (Figure1Ba). Meanwhile, PPC8/PTX@Ma maintained its globularshape intact with changing pH (Figure 1Bb). DLS data further

Figure 2. (A) Illustration of membrane escape and drug-release mechanism. Cumulative drug-release profile of PPiP/PTX@Ma (B) and PPC8/PTX@Ma (C) at various pH values. DLS (D, E) and TEM (F, G) images of PPiP/PTX in buffer at pH 7.4 and 6.5. Changes in the ζ potential (H)and CMC value (I) of the polymers PPiP and PPC8 at various pH values.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1910

confirmed the change in the size of membrane-coatednanoparticles with hydration status (Figure S7A,B). Meanwhile,both formulations exhibited a small negative charge with themembrane coating in comparison with that of the nudeparticles (Figure S7C). This interesting phenomenon suggestedthat the change in morphology might provide a way to achievemembrane decortication and discharge of the interior nano-particle in response to the acidity of tumor tissue. The releasekinetics of both formulations were then investigated in milieus

simulating tumor tissue and intracellular acidity (Figure 2A).When the pH was decreased from 7.4 to 6.5, PPiP/PTX@Mashowed only a slight rise in the accumulative release plateau(Figure 2B), although the external membrane was cracked(Figure 1B). Meanwhile, the internal drug-loading nano-particles (PPiP/PTX) maintained their nanoscale shape intactdespite morphological expansion from pH 7.4 (Figure 2D,F) topH 6.5 (Figure 2E,G). When the pH approached 5.0,simulating the endosomal pH environment, the drug release

Figure 3. (A) Penetration efficiency of PPC8@Ma and cskc-PPiP@Ma into tumor spheroids with 2 h of incubation. Upper panels show results inpH 7.4 medium and lower panels in pH 6.5, scale bar 100 μm, z-axis depth 20 μm. (B) Tumor sections of mice 24 h after injection with cskc-PPiP,PPC8@Ma, PPiP@Ma, and cskc-PPiP@Ma, scale bar 100 μm. DAPI (blue), CD34 (green), and Probe (red) stained the cell nuclei, blood vessels,and formulation trace, respectively.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1911

of PPiP/PTX@Ma was rapidly enhanced by the additional H+.The release behavior of PPiP/PTX (Figure S8A) was consistentwith that of PPiP/PTX@Ma, which confirmed that the drug

unpacking relied on particle disassembly rather than themembrane escape process. In contrast, PPC8/PTX@Mashowed consistent size and appearance at pH 7.4 and 6.5

Figure 4. Biodistribution and in vivo antitumor efficacy. (A) IVIS images of mice injected with near-infrared probe-loaded cskc-PPiP and cskc-PPiP@Ma at designated time points. (B) 3D reconstruction of fluorescence signal in cskc-PPiP@Ma-treated mouse at 48 h. (C) Heart (H), liver (Li),spleen (S), lung (Lu), kidney (K), and tumor (T) excised from the above-mentioned mouse. (D) Quantification of PTX concentration in organs andtumor tissue excised from mice treated with Taxol, cskc-PPiP/PTX, PPC8/PTX@Ma, and cskc-PPiP/PTX@Ma (n = 4). Body weight (E) andtumor volume (F) data were recorded during the 3 week treatment course. (G) Tumor tissue apoptosis in mice treated with saline, Taxol, PPC8/PTX@Ma, PPiP/PTX@Ma, and cskc-PPiP/PTX@Ma. Green signals indicate apoptotic cells in tumor section. Scale bar, 100 μm.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1912

(Figure S9), which was consistent with its slow release at all pHvalues (Figure 2C), even slower than that of PPC8/PTX(Figure S8B). This result indicated that the membrane-coatingstrategy contributed to the integrity and stability of theformulation. Changes in the ζ potential and critical micelleconcentration (CMC) values of different polymeric nano-particles indicated that the stepwise protonation of the PPiPpolymer might contribute to the morphological expansion andgradual loss of drug-loading capacity of PPiP/PTX@Ma(Figure 2H,I). A titration test showed good buffering capacityof PPiP in the pH range 6.0−7.0, while it became too cationicto complete micellization at pH 5.0 (Figure S10). In the acidicextracellular microenvironment, the interior PPiP/PTX nano-particle performed as a H+-absorbing proton sponge and couldescape from the ruptured coating after equilibrium disruptionof the capsular membrane structure. After uptake by the tumorcells, the even lower intracellular pH finally caused nanoparticledisassembly and drug release. The buffering capacity derivedfrom the chemical structure of the polymeric materials led todifferent drug-release behavior under different biomimeticconditions.The cellular internalization behavior of cskc-PPiP/PTX was

then tested in the MDA-MB-231 cell line. A syntheticoligopeptide targeting IGF1R was chosen to facilitate uptakeby breast-cancer cells.29,30 The ligand modification rate wasoptimized to 20%, as increased modification did not furtherimprove the uptake efficiency (Figure S11A,B). The internal-ization mechanism of cskc-PPiP/PTX was then studied byindividually inhibiting different endocytosis pathways bytreating the cells with filipin complex (caveolae-mediatedpathway), phenylarsine oxide (PhAsO and clathrin-dependentpathways), colchicine (macropinocytosis), and ice incubation(ATP-dependent pathway). The results showed that the uptakeof coumarin-loaded particles (cskc-PPiP/coumarin) mainlyrelied on ATP-facilitated caveolae-mediated and clathrin-dependent pathways (Figure S11C,D). Confocal images ofcskc-PPiP/coumarin traces in the cytoplasm showed obviousoverlap with the acidic late endosome after 0.5 h of incubation,suggesting that cskc-PPiP/PTX might follow the sameinternalization process and undergo second-stage intracellulardrug release (Figure S11E).Based on the in vitro uptake results in monolayer tumor cells,

we further explored the tumor-penetrating ability of cskc-PPiP@Ma and the control PPC8@Ma both in vitro and in vivoby constructing red BODIPY-loaded formulations (Figure 3A).We conducted an in vitro experiment on tumor spheroids attwo different pH values to evaluate the correlation with internalnanoparticle escape. In the blood-mimicking milieu (pH 7.4),both cskc-PPiP@Ma and PPC8@Ma showed only surfaceadsorption on tumor spheroids because their size limitedinfiltration into tight intercellular junctions without additionalassistance. In the tumor-tissue microenvironment-mimickingmilieu (pH 6.5), cskc-PPiP@Ma displayed considerablepenetration efficiency since the internal nanoparticles under-went membrane escape. Meanwhile, PPC8@Ma maintainedmembrane integration and adsorption on the spheroid surface.In addition to the size and flexibility advantages resulting frommembrane escape, the targeting ligand was also found tocontribute to the general penetration efficiency. The pene-tration data of PPiP@Ma and cskc-PPiP@Ma demonstratedthat cskc modification facilitated faster penetration than that ofplain PPiP after membrane escape (Figure S12, pH 6.5, 0.75 hof incubation). The penetration study was then repeated in

mouse models bearing orthotopic tumors. Tumor sections werestained with DAPI (blue) to locate tumor cells and CD34(green) to show intratumor vessels (Figure 3B). Anotherfluorescence probe (red) allowed tracing of the internalnanoparticles in the tumor tissue. Significantly greater tumoraccumulation and distal penetration from the vessel toward theinterior of the tumor were found in the PPiP@Ma- and cskc-PPiP@Ma-treated groups than in the PPC8@Ma-treatedgroup. Moreover, cskc-PPiP@Ma displayed generally higherfluorescence intensity than PPiP@Ma, indicating greatercellular uptake, which was deduced to be related to theIGF1R-mediated uptake pathway.Furthermore, general biodistribution was investigated by the

injection of near-infrared fluorescence-probe-loaded cskc-PPiP@Ma and cskc-PPiP followed by observation at designatedtime points (2, 4, 12, 24, and 48 h; Figure 4A). The cskc-PPiP@Ma group showed rapid tumor homing within 2 h andlong residence at the cancerous site with only slight signalattenuation until 48 h, whereas the cskc-PPiP group displayedsignificant liver and kidney accumulation in the first 2 h andgradual clearance in 24 h. The 3D reconstructed fluorescentphotograph of the cskc-PPiP@Ma-treated mouse also demon-strated the favorable tumor accumulation of PPiP@Ma (Figure4B). The major organs (heart, liver, spleen, lung, and kidney)and whole tumor tissue showed similar distribution tendencies(Figure 4C, left cskc-PPiP, right cskc-PPiP@Ma). Thedistribution profiles of various drug-loading formulations werefurther evaluated by the quantitative determination of the PTXconcentrations in perfused and saline-washed organs and tumortissues 24 h postinjection (Figure 4D and Figure S13). Becauseof its tumor-tissue pH-responsive escape mechanism and activetargeting, cskc-PPiP/PTX@Ma showed greater tumor accumu-lation than PPC8/PTX@Ma. In addition, despite its lack ofmembrane coating, cskc-PPiP/PTX showed greater tumoraccumulation than PTX in 50% cremophor−ethanol solutiondue to its active targeting ability.Since slight fluorescent signals and drug accumulation were

observed in other organs, we assessed the systemic toxicity ofvarious formulations. The polymers (PPiP, PPC8) andbiologically derived membrane materials (Ma) were found tobe nontoxic to 293 cells, which indicated their goodbiocompatibility and low toxicity (Figure S14A). The uptakeexperiment in the 293 cell line also showed little PTXinternalization, as the intact membrane coating protected theinternal drug-loading nanoparticles (Figure S14B). H&E-stained sections showed inflammation in the portal area ofthe liver treated with Taxol, while no significant organic injurywas observed in organs from the mice treated with membrane-coated formulations (Figure S14C). The inflammation mayhave been caused by the cosolvents of Taxol, ethanol, andcremophor, which could induce liver damage with repeatedinjection.31,32

In the next pharmacodynamics experiments, MTT andapoptosis tests were conducted to evaluate the in vitrocytotoxicity of the naked drug-loaded nanoparticles to MDA-MB-231 cells. The cell viability and IC50 values demonstratedthat all the naked particle formulations exhibited validantitumor efficacy in vitro (Figure S15A). To further elucidatethe antitumor efficacy of those membrane-coated nanoparticles,we evaluated the apoptosis of MDA-MB-231 cells treated withTaxol, PPC8/PTX@Ma, PPiP/PTX@Ma, and cskc-PPiP/PTX@Ma, all in PBS pH 6.5 before incubation. The annexin-V (green) signal indicated everted phosphatidylserine during

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1913

early apoptosis, while PI (red) marked membrane-permeablecells in late apoptosis (Figure S15B). The MTT results showedthat PPC8/PTX@Ma, because the drug-loaded nanoparticleswere sealed in intact membrane coatings, had lower cytotoxicitythan PPiP/PTX@Ma and cskc-PPiP/PTX@Ma, as theenclosed drug-loading nanoparticles escaped from the mem-brane vesicles during PBS incubation via the proton spongeeffect, unleashing their antitumor efficacy. The cskc-PPiP/PTX@Ma demonstrated particularly enhanced antitumorability because of its facilitated cellular internalization. The invivo antitumor effects of various PTX formulations were furtherinvestigated in an orthotopic breast-cancer tumor model byintravenous administration every 4 days for 3 weeks. The mice’sbody weight and tumor volume were recorded every 2 days toevaluate the general toxicity and antitumor efficacy (Figure4E,F and Figure S16). The cskc-PPiP/PTX@Ma-treated groupshowed significant control of the tumor burden whilemaintaining a healthy body weight. The mice in the Taxolgroup suffered from continuous weight loss, mainly due tononselective biodistribution and excipient toxicity (FigureS14C). Tumor sections from mice in each group were stainedfor tissue-level apoptosis detection (Figure 4G). The samplesfrom the cskc-PPiP/PTX@Ma group displayed the mostextensive cell apoptosis, indicating the remarkable in vivoantitumor effect of cskc-PPiP/PTX@Ma.In summary, we have developed a macrophage-membrane-

coated nanoparticle delivery system with a step-by-step releaseprofile in response to the pH differences in the tumormicroenvironment. The resulting formulation (cskc-PPiP/PTX@Ma) exhibited favorable tumor-homing ability insystemic circulation and high biocompatibility resulting fromits membrane coating. The unique buffering property of thePPiP materials for the interior nanoparticle further endowedthe formulation with tuned drug-release kinetics responsive toextracellular and intracellular tumor microenvironment stimuli.The combination of biomimetic cell membranes and responsivepolymeric nanoparticles could inspire the rational design ofmembrane-coating systems for tailored chemo-drug delivery totumors.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.nano-lett.7b05263.

Detailed methods, synthesis and characterization ofpolymers, characterization and pharmacodynamics dataof control formulations, and internalization mechanismstudy (PDF)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: jianjunc@illinois.edu.*E-mail: jiangchen@shmu.edu.cn.ORCIDKaimin Cai: 0000-0001-9442-8312Yongzhuo Huang: 0000-0001-7067-8915Jianjun Cheng: 0000-0003-2561-9291Chen Jiang: 0000-0002-4833-9121NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We acknowledge the support from National Science Fund forDistinguished Young Scholars (Grant 81425023), NationalNatural Science Foundation of China (Grant 81373355), andFudan-SIMM Joint Research Fund (Grant FU-SIMM20174009). Jianjun Cheng acknowledges support from theNSF (DMR-1309525) and NIH (R01 1R01CA207584 andR21 CA198684). Kaimin Cai acknowledges Beckman InstituteGraduate Fellowship support at the University of Illinois atUrbana−Champaign. Yu Zhang acknowledges China Scholar-ship Council (CSC).

■ REFERENCES(1) Cai, K.; Wang, A. Z.; Yin, L.; Cheng, J. J. Controlled Release 2017,263, 211−222.(2) Ryan, S. M.; Brayden, D. J. Curr. Opin. Pharmacol. 2014, 18,120−8.(3) Liu, T.; Choi, H.; Zhou, R.; Chen, I. W. PLoS One 2014, 9,e103576.(4) Magana, I. B.; Yendluri, R. B.; Adhikari, P.; Goodrich, G. P.;Schwartz, J. A.; Sherer, E. A.; O’Neal, D. P. Ther. Delivery 2015, 6,777−83.(5) Knop, K.; Hoogenboom, R.; Fischer, D.; Schubert, U. S. Angew.Chem., Int. Ed. 2010, 49, 6288−308.(6) Jiang, S.; Cao, Z. Adv. Mater. 2010, 22, 920−32.(7) Hu, C. M.; Zhang, L.; Aryal, S.; Cheung, C.; Fang, R. H.; Zhang,L. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 10980−5.(8) Luk, B. T.; Zhang, L. J. Controlled Release 2015, 220, 600−7.(9) Hu, C. M.; Fang, R. H.; Copp, J.; Luk, B. T.; Zhang, L. Nat.Nanotechnol. 2013, 8, 336−40.(10) Gao, W.; Hu, C. M.; Fang, R. H.; Luk, B. T.; Su, J.; Zhang, L.Adv. Mater. 2013, 25, 3549−53.(11) Rao, L.; He, Z.; Meng, Q. F.; Zhou, Z.; Bu, L. L.; Guo, S. S.; Liu,W.; Zhao, X. Z. J. Biomed. Mater. Res., Part A 2017, 105, 521−530.(12) Xuan, M.; Shao, J.; Dai, L.; He, Q.; Li, J. Adv. Healthcare Mater.2015, 4, 1645−52.(13) Baek, S. K.; Makkouk, A. R.; Krasieva, T.; Sun, C. H.; Madsen, S.J.; Hirschberg, H. J. Neuro-Oncol. 2011, 104, 439−48.(14) Madsen, S. J.; Christie, C.; Hong, S. J.; Trinidad, A.; Peng, Q.;Uzal, F. A.; Hirschberg, H. Lasers Med. Sci. 2015, 30, 1357−65.(15) Long, K. B.; Beatty, G. L. Oncoimmunology 2013, 2, e26860.(16) Hagemann, T.; Balkwill, F.; Lawrence, T. Cancer Cell 2007, 12,300−1.(17) Hyvonen, Z.; Hamalainen, V.; Ruponen, M.; Lucas, B.; Rejman,J.; Vercauteren, D.; Demeester, J.; De Smedt, S.; Braeckmans, K. J.Controlled Release 2012, 162, 167−75.(18) Rehman, Z. u.; Hoekstra, D.; Zuhorn, I. S. ACS Nano 2013, 7,3767−3777.(19) Neuberg, P.; Kichler, A. Adv. Genet. 2014, 88, 263−88.(20) Shao, K.; Zhang, Y.; Ding, N.; Huang, S.; Wu, J.; Li, J.; Yang, C.;Leng, Q.; Ye, L.; Lou, J.; Zhu, L.; Jiang, C. Adv. Healthcare Mater.2015, 4, 291−300.(21) Liu, Y.; Li, J.; Shao, K.; Huang, R.; Ye, L.; Lou, J.; Jiang, C.Biomaterials 2010, 31, 5246−57.(22) Zheng, N.; Song, Z.; Liu, Y.; Zhang, R.; Zhang, R.; Yao, C.;Uckun, F. M.; Yin, L.; Cheng, J. J. Controlled Release 2015, 205, 231−9.(23) Yoneda, T.; Hiasa, M.; Nagata, Y.; Okui, T.; White, F. Biochim.Biophys. Acta, Biomembr. 2015, 1848, 2677−84.(24) Peppicelli, S.; Bianchini, F.; Calorini, L. Cancer Metastasis Rev.2014, 33, 823−32.(25) Anderson, D. G.; Lynn, D. M.; Langer, R. Angew. Chem., Int. Ed.2003, 42, 3153−8.(26) Anderson, D. G.; Akinc, A.; Hossain, N.; Langer, R. Mol. Ther.2005, 11, 426−34.(27) Tian, X.; Aruva, M. R.; Qin, W.; Zhu, W.; Duffy, K. T.; Sauter,E. R.; Thakur, M. L.; Wickstrom, E. J. Nucl. Med. 2004, 45 (12),2070−2082.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1914

(28) Tian, X.; Chakrabarti, A.; Amirkhanov, N. V.; Aruva, M. R.;Zhang, K.; Mathew, B.; Cardi, C.; Qin, W.; Sauter, E. R.; Thakur, M.L.; Wickstrom, E. Ann. N. Y. Acad. Sci. 2005, 1059, 106−44.(29) Litzenburger, B. C.; Creighton, C. J.; Tsimelzon, A.; Chan, B.T.; Hilsenbeck, S. G.; Wang, T.; Carboni, J. M.; Gottardis, M. M.;Huang, F.; Chang, J. C.; Lewis, M. T.; Rimawi, M. F.; Lee, A. V. Clin.Cancer Res. 2011, 17, 2314−27.(30) Sun, B. F.; Kobayashi, H.; Le, N.; Yoo, T. M.; Drumm, D.; Paik,C. H.; McAfee, J. G.; Carrasquillo, J. A. Cancer Res. 1997, 57, 2754−9.(31) Liebmann, J. E.; Cook, J. A.; Lipschultz, C.; Teague, D.; Fisher,J.; Mitchell, J. B. Br. J. Cancer 1993, 68, 1104−9.(32) Liebmann, J.; Cook, J. A.; Mitchell, J. B. Lancet 1993, 342, 1428.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.7b05263Nano Lett. 2018, 18, 1908−1915

1915