Biomaterials 35 (2014) 5393e5406

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Polyglycerol-coated nanodiamond as a macrophage-evading platform for selective drug delivery in cancer cells Li Zhao a, Yong-Hong Xu b, Tsukasa Akasaka c, Shigeaki Abe c, Naoki Komatsu a, Fumio Watari c, Xiao Chen c, * a b c

Department of Chemistry, Shiga University of Medical Science, Seta, Otsu 520-2192, Japan Institute of Ophthalmological Research, Department of Ophthalmology, Renmin Hospital of Wuhan University, 430060 Wuhan, China Department of Biomedical, Dental Materials and Engineering, Graduate School of Dental Medicine, Hokkaido University, Sapporo 060-8586, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 January 2014 Accepted 17 March 2014 Available online 8 April 2014

A successful targeted drug delivery device for cancer chemotherapy should ideally be able to avoid nonspecific uptake by nonmalignant cells, particularly the scavenging monocyte-macrophage system as well as targeting efficacy to bring the drug preferentially into tumor cells. To this purpose, we developed a platform based on detonation nanodiamond (dND) with hyperbranched polyglycerol (PG) coating (dNDPG). dND-PG was first demonstrated to evade non-specific cell uptake, particularly by macrophages (U937). RGD targeting peptide was then conjugated to dND-PG through multistep organic transformations to yield dND-PG-RGD that still evaded macrophage uptake but was preferentially taken up by targeted A549 cancer cells (expressing RGD peptide receptors). dND-PG and dND-PG-RGD showed good aqueous solubility and cytocompatibitlity. Subsequently, the anticancer agent doxorubicin (DOX) was loaded through acid-labile hydrazone linkage to yield dND-PG-DOX and dND-PG-RGD-DOX. Their cellular uptake and cytotoxicity were compared against DOX in A549 cells and U937 macrophages. It was found that dNDPG-DOX uptake was substantially reduced, displaying little toxicity in either type of cells by virtue of PG coating, whereas dND-PG-RGD-DOX exerted selective toxicity to A549 cells over U937 macrophages that are otherwise highly sensitive to DOX. Finally, dND-PG was demonstrated to have little influence on U937 macrophage cell functions, except for a slight increase of TNF-a production in resting U937 macrophages. dND-PG is a promising drug carrier for realization of highly selective drug delivery in tumor cells through specific uptake mechanisms, with minimum uptake in and influence on macrophages. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Polyglycerol Coating Detonation nanodiamond RGD peptide Selective delivery Cancer cells

1. Introduction In drug treatment of cancerous diseases, systemically administered chemotherapeutic agents are distributed throughout the body acting indiscriminately on both cancer and normal tissues. This often leads to inadequate therapeutic efficacy due to insufficient drug availability in cancer tissues and significant toxicity in normal tissues. Thus, preferential accumulation of chemotherapeutic drugs in the malignant cells is highly desirable and carrierbased, highly targeted drug delivery systems (DDS) are instrumental in realization of this strategy. While the targeting moiety largely determines the specificity and efficacy of a DDS for cancer therapy, it is recognized that the success of a targeted DDS also

* Corresponding author. Tel.: þ81 11 706 4250; fax: þ81 11 706 4251. E-mail address: [email protected] (X. Chen). http://dx.doi.org/10.1016/j.biomaterials.2014.03.041 0142-9612/Ó 2014 Elsevier Ltd. All rights reserved.

depends on how well it avoids the non-specific uptake by nonmalignant cells and the scavenging system i.e. the mononuclear phagocyte system (MPS). The MPS consists of specialized phagocytic cells including monocytes and macrophages in the reticular connective tissue, Kupffer cells of the liver and tissue histiocytes [1]. MPS is a vital component of the immune system performing crucial immune functions i.e. antigen presentation, phagocytosis, and immunomodulation through production of cytokines and growth factors [1]. Recognition and elimination of exogenous particles that enters the body is a primary function of the MPS. Elimination by the MPS negatively affects the in vivo efficacy and circulation time of a particle-based DDS [2,3]. MPS can also be activated by an exogenous particulate matter to initiate immune responses [4,5]. Various targeted DDS based on nanoparticles have thus far been devised that focus on increasing accumulation of chemotherapeutic drugs in the malignant cells [6,7], but few have addressed the issue of scavenging by the MPS and their influence on the MPS.

5394

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

Fig. 1. STEM images of (A) dND, (B) dND-PG, (C) dND-PG-DOX and (D) dND-PG-RGD-DOX. All scale bars represent 100 nm.

Non-specific cellular uptake of particulate matter is largely dependent on the surface chemistry and surface charge of the particles [8,9]. Thus, drug carrier particles are often coated with an electrically neutral hydrophilic surface layer, a coating that shields against non-specific cellular uptake [9]. The most widely used coating material with this effect thus far is polyethylene glycol (PEG), a non-ionic hydrophilic polymer that prevents the adsorption of proteins and subsequent cellular uptake. Despite its wide usage, such significant drawbacks have been found of PEG coating as impeding endosomal escape [10] and inducing immune response [11]. A variety of alternative coating materials have been

proposed, among which is hyperbranched polyglycerol (PG) [9,12]. Hyperbranched PG has recently attracted attention for biomedical application due to its biocompatibility, hydrophilic property and anti-fouling effect [13e16]. It was reported that PG polymers have long plasma half-lives [17] and PG coating can prolong liposome circulation [16,17]. In addition, PG is more amenable for further functionalization than PEG due to numerous surface hydroxyl groups [18,19]. Nanodiamond (ND) is a promising platform for biomedical applications such as imaging and drug/gene delivery due to its high specific surface area, tunable surface structures and

Fig. 2. FTIR spectra of (A) dND-PG, (B) dND-PG-OTs, (C) dND-PG-N3, (D) dND-PG-N3-PhNO2, (E) dND-PG-N3-NHNH2 and (F) dND-PG-RGD-NHNH2. Arrows indicate new absorption bands in each step.

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

5395

Scheme 1. Synthesis of dND-PG-RGD from dND. i) glycidol, 140  C, 20 h; ii) p-TsCl, NaOH, 0  C w r. t., overnight; iii) NaN3, 90  C, overnight; iv) RGD propargyl amide, copper (II) sulfate pentahydrate, sodium ascorbate, r. t., 48 h.

biocompatibility [20e25]. Moreover, nitrogen-vacancy (N-V) center in ND emits non-bleaching and non-blinking fluorescence at 550e800 nm, which enables in vitro fluorescence cell labeling and in vivo imaging [26e28]. In a previous work [29], we have developed a drug carrier based on ND with a surface coating of PG (NDPG). To achieve active tumor cell targeting, ND-PG was conjugated with cyclic ArgeGlyeAsp (RGD) peptide, which specifically binds to integrin receptor avb3 that is over-expressed in multiple types of

malignant tumor and tumor tissue cells [30,31], yielding the targeted carrier ND-PG-RGD. For fluorescent tracking, red emissive fND-PG-RGD was constructed with a fluorescent ND (fND) core having a mean diameter of ca. 50 nm. Using in vitro cell models, we demonstrated that the PG coating masks fND-PG-RGD from nonspecific uptake by non-phagocytic cells but permits selective uptake by tumor cells over-expressing integrin receptor avb3. Consequently, highly preferential drug delivery and selective killing of

Scheme 2. Synthesis of dND-PG-RGD-DOX from dND-PG-N3. i) bis(4-nitrophenyl) carbonate, triethylamine, r. t., 24 h; ii) hydrazine monohydrate, 90  C, overnight; iii) RGD propargyl amide, copper (II) sulfate pentahydrate, sodium ascorbate, r. t., 48 h; iv) doxorubicin hydrochloride, pH 7, 50  C, 24 h.

5396

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

malignant cells was achieved when a platinum-based drug was loaded onto ND-PG-RGD and applied to the cell models. In the current work, we first refined the ND-PG/ND-PG-RGD platform using a detonation ND (dND) core with a diameter of 4e5 nm, resulting in dND-PG/dND-PG-RGD which then received physical and chemical characterization. The main purposes of this study are to investigate 1) capability of dND-PG/dND-PGRGD to evade macrophage uptake, 2) utility of this property in achieving selective drug delivery and toxicity in cancer cells sparing macrophages, and 3) the influence of dND-PG/dND-PGRGD on cell functions of macrophage. In vitro human cancers cells and macrophages were employed. For study of selective drug delivery and toxicity, the anticancer agent doxorubicin (DOX) was loaded to give dND-PG-DOX and dND-PG-RGD-DOX. Results of this study highlight the property of dND-PG/dNDPG-RGD against macrophage uptake and its potential usefulness in cancer treatment.

characterized by scanning transmission electron microscopy (STEM) (Fig. 1B), FTIR (Fig. 2A) and 1H NMR (Fig. S1A). 2.1.2. Immobilization of RGD peptide and DOX on dND-PG The grafted PG not only largely increases the aqueous dispersibility of dND-PG (>80 mg/mL in water), but also provides numerous amendable hydroxyl groups for further functionalization [13,32,33]. The RGD peptide was covalently immobilized on dND-PG to give dND-PG-RGD through click reaction according to our recently reported method (Scheme 1) [33]. Briefly, some of the hydroxyl groups of dND-PG were first reacted with tosyl chloride (TsCl) and the resultant tosylates (dND-PG-OTs) were substituted by azido groups (dND-PG-N3). dND-PG-N3 was further conjugated with RGD peptide through click chemistry of the azido groups with the propiolic amide of RGD peptide, yielding dND-PG-RGD. To immobilize DOX on dND-PG-RGD, hydrazine moiety was introduced through substitution of the p-nitrophenyl group in the p-nitrobenzoate (dND-PG-N3-PhNO2) to give dND-PG-N3-NHNH2 as shown in Scheme 2. After click conjugation of RGD peptide (dND-PG-RGD-NHNH2), doxorubicin was immobilized as hydrazone (dNDPG-RGD-DOX). dND-PG-DOX without RGD peptide was also prepared as a control from dND-PG-NHNH2. The experimental details of the above transformations are shown in Supplementary Data. 2.2. In vitro drug release

2. Materials and methods 2.1. Preparation and characterization of dND-PG and derivatives 2.1.1. Synthesis of dND-PG dND-PG was prepared according to our reported method [13,32], using dND particles with a diameter of 4e5 nm (Fig. 1A) as a starting material. It was

dND-PG-DOX (3.0 mg) was well dispersed in 24 mL of phosphate-citrate buffer (pH 5.0) and phosphate buffer solution (PBS, pH 7.4), respectively. The solution was equally allotted to 8 glass tubes and sealed. All tubes were then shaken at 37  C at 200 rpm (TAITEC Bio-Shaker, BR-30L). Each type of buffer in 2 tubes was taken out at predetermined time intervals for analysis. The solid was precipitated by ultracentrifugation and the resulting supernatant was collected for UVeVis absorption

Fig. 3. (A) UVeVis absorption of DOX-HCl, dND-PG and dND-PG-DOX in water. The inset shows photographs water solutions of DOX-HCl, dND-PG and dND-PG-DOX. (B) PL spectra of DOX-HCl and dND-PG-DOX in water excited at 480 nm. (C) FTIR spectra of DOX-HCl, dND-PG and dND-PG-DOX. (D)In vitro drug release profiles of dND-PG-DOX at pH 5.0 and pH 7.4.

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406 measurement. Concentration of released DOX was calculated using the absorbance at 480 nm according to a linear standard curve. 2.3. Cell culture Human leukemic monocyte lymphoma cells (U937) and lung adenocarcinoma epithelial cells (A549) were purchased from Japanese Collection of Research Bioresources Cell Bank and cultured with RPMI-1640 medium (Sigma Aldrich, USA) supplemented with 10% fetal bovine serum (Sigma Aldrich) in a humidified incubator (5% CO2/95% air atmosphere at 37  C). U937 cells were differentiated into macrophages through incubation with culture medium containing 100 ng/mL of phorbol 12-myristate 13-acetate (PMA, Sigma Aldrich) for 48 h [34]. The PMAtreated cells were then washed with PBS to remove PMA and non-adherent cells. The adherent cells, referred to as U937 macrophages, were subjected to treatment of ND-based materials following recovering in fresh culture medium for 24 h. To activate U937 macrophages, the cells were incubated with culture medium containing 5 mg/mL of lipopolysaccharides (LPS, Sigma Aldrich, USA) for 24 h. 2.4. Cytocompatibility assay A549 cells and U937 macrophages cultured in 96-well plates were treated with dND-PG and dND-PG-RGD at concentrations up to 400 mg/mL, 100 mL per well for 24 h. dND-PG and dND-PG-RGD were then removed and cell viability was assayed using CCK-8 kit as described in the manual provided by the kit manufacture (Dojindo Molecular Technologies, Inc., Japan). 2.5. Cell uptake analysis For cell uptake analysis of dND-PG and dND-PG-RGD, A549 cells and U937 macrophages cultured in 24-well plates were treated with each material dispersed in culture medium at concentrations up to 200 mg/mL, 500 mL per well for 24 h. For cell uptake analysis of dND-PG-DOX, dND-PG-RGD-DOX and DOX, A549 cells and U937 macrophages cultured in 24-well plates were incubated with each material in culture medium, 500 mL per well for 24 h. Concentrations are normalized to DOX up to 4 mg/mL. Cells in each well were then lifted by trypsin and analyzed by flow cytometry (FACS) on a FACSCalibur flow cytometer (BD, Bioscience). At least ten thousand cells were measured per well. Side scatter (SSC) intensity on the arithmetic scale was acquired for uptake analysis of dND-PG and dND-PG-RGD and red fluorescence (FL3-H) intensity acquired for uptake analysis of dND-PG-DOX, dNDPG-RGD-DOX and DOX. Data were processed with CellQuest (BD) and WinMDI 2.9 software. Arithmetic means were used for quantification of uptake.

5397

2.9.1. Phagocytosis U937 phagocytic function was assayed using carboxylate modified polystyrene yellow-green fluorescent latex beads (2.5%, diameter 1 mm, Sigma Aldrich). Briefly, the latex beads were washed with distilled water, centrifuged at 10,000 g for 8 min at room temperature and incubated in 3% BSA containing 25 mM Na3PO4 (pH 6.0) for 15 min with bath sonication. The beads were then washed once with culture medium containing 10% FBS and re-suspended in the same culture medium at 2.0% concentration. This is latex beads stock suspension. Working suspension was prepared by diluting the stock suspension to 0.03% concentration with culture medium containing 10% FBS and applied to each well of cells, 0.5 mL per well. After incubation in the dark at 37  C for 90 min, the latex beads were removed. Excess beads were removed by washing 3 times with PBS. The cells in each well were then lifted by trypsin and analyzed by FACS for cellular fluorescence indicative of phagocytosis of latex beads. Percentage of phagocytic cells (with high green fluorescence) was calculated. 2.9.2. Reactive oxygen species (ROS) generation Both resting and LPS-activated U937 macrophages were incubated with culture medium containing 20 mM of 20 ,70 -dichlorofluorescein diacetate (DCFH-DA, Invitrogen), 1 mL per well in the dark at 37oC for 30 min. The cells were then detached and analyzed by FACS for cellular green fluorescence indicative of ROS level. Geometric means were used for quantification. 2.9.3. TNF-a secretion TNF-a secretion in the supernatant by U937 macrophages, either LPS-activated or not, was measured using a Human TNF-a ELISA Kit (Invitrogen, KHC 3011) as described in the manual provided by the kit manufacturer.

3. Results and discussion The introduction of functional groups on dND-PG through the organic transformations was confirmed by FTIR (Fig. 2). The absorption bands at 1597, 1350 and 1176 cm1 are attributed to C]C stretching of the benzene ring, asymmetric and symmetric

2.6. Cytotoxicity assay A549 cells and U937 macrophages cultured in 24-well plates were treated with dND-PG-DOX, dND-PG-RGD-DOX and DOX in culture medium, 500 mL per well. Concentrations were normalized to DOX up to 4 mg/mL. After incubation for 24 h or 48 h, dND-PG-DOX, dND-PG-RGD-DOX and DOX were removed and cell toxicity was then assayed using CCK-8 kit as described in the manual provided by the kit manufacture (Dojindo Molecular Technologies, Inc., Japan). 2.7. Fluorescent microscopy of mixed cell culture Mixed culture of A549 cells and U937 macrophages were first performed. Briefly, A549 cells were labeled with 2 mM of 5(6)-carboxyfluorescein diacetate, succinimidyl ester (CFDA-SE) [35] and seeded in 24-well plates containing U937 macrophages and then cultured for 24 h. The mixed cell culture was then incubated respectively with dND-PG-DOX, dND-PG-RGD-DOX and DOX in culture medium, 500 mL per well. Concentrations were normalized to 2 mg/mL of DOX. The incubation was stopped after 30 min, 24 h and 48 h. The cells were washed with PBS, fixed with 4% paraformaldehyde and then observed with an IX70 inverted microscope (Olympus, Tokyo, Japan). 2.8. Fluorescence microscopy colocalization analysis A549 cells grown in 24-well plates were treated with dND-PG-DOX, dND-PGRGD-DOX and DOX in culture medium, 500 mL per well for 24 h. Concentrations were normalized to 2 mg/mL of DOX. Thereafter, dND-PG-DOX, dND-PG-RGD-DOX and DOX was removed by PBS washing and the cells were stained with 100 nM LysoTrackerÒ Blue (Invitrogen) in culture medium, 0.5 mL per well for 1 h. Fluorescent microphotographs were then taken using an IX70 inverted microscope (Olympus, Tokyo, Japan). 2.9. Macrophage cell function analysis U937 macrophages cultured in 24-well plates were treated with dND-PG and dND-PG-RGD at concentrations up to 200 mg/mL, 500 mL per well for 24 h dND-PG and dND-PG-RGD were then removed and the cells were washed twice with PBS. Thereafter, the cells were either analyzed as described below or activated with LPS as described in the cell culture section. LPS-activated U937 macrophages then received the same analyses as follows.

Fig. 4. Effects of dND-PG and dND-PG-RGD on cell viability assayed by CCK-8 reduction. Duration of treatment was 24 h. Values are means  SD (n ¼ 6). Student’s t-test was performed.

5398

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

stretchings of S/O of the tosyl group in dND-PG-OTs, respectively (Fig. 2B). The dND-PG-N3 clearly shows a characteristic absorption band at 2100 cm1 corresponding to azido group (Fig. 2C). The strong absorption bands at 1790 and 1600 cm1 are assigned to C]O stretching of carbonate and aromatic C]C stretching in dND-PG-N3-PhNO2. The characteristic stretching bands of eNO2 are detected at 1450 and 1265 cm1 (Fig. 2D). The intense absorption band at 1690 cm1 is attributed to NeH bending of hydrazine groups in dND-PG-N3-NHNH2 (Fig. 2E). The azido absorption band disappeared after the click conjugation of the targeting peptide (Fig. 2F), indicating the complete consumption of azido groups. The immobilization of RGD peptide was verified by the absorption bands at 1650 and 1590 cm1, which correspond to the C]O stretching and NeH bending of amide bonds in the peptide. Although the reddish color (in web version) of the dND-PG-DOX solution indicated the presence of DOX (Fig. 3A, inset), dND-PGDOX was further characterized by UVeVis absorption, photoluminescence (PL) and FTIR to confirm the drug loading. The UV-Vis absorption spectrum of DOX hydrochloride (DOX-HCl), dND-PG, and dND-PG-DOX are shown in Fig. 3A. dND-PG-DOX shows the same absorption band as DOX-HCl at 480 nm, supporting that DOX moiety is connected with dND-PG as shown in Scheme 1. The characteristic absorbance was used to quantify the amount of the loaded DOX on dND-PG. Fig. 3B shows the PL spectra of the aqueous solutions of dND-PG-DOX and DOX-HCl. The emission peaks of both solutions were observed at 560 nm. FTIR spectrum of dND-PG-DOX reveals a characteristic C]C stretching at 1580 cm1 corresponding to the aromatic ring of DOX moiety (Fig. 3C).

Fig. 5. Cell uptake of dND-PG and dND-PG-RGD following 24-hour treatment. FACS side scatter (SSC) intensity was acquired on arithmetic scale and arithmetic mean was used for quantification. Values are means  SD (n ¼ 6). Student’s t-test was performed (** and ##p < 0.01).

The aqueous solubility and stability of the dND-PG-DOX and dND-PG-RGD-DOX were tested in water and PBS. Both dND-PGDOX and dND-PG-RGD-DOX showed excellent solubility in both water and PBS with concentration of >1.0 mg/mL. The size of dNDPG-DOX and dND-PG-RGD-DOX was determined by dynamic light scattering (DLS) and STEM. The mean hydrodynamic diameter of dND-PG-DOX (49.7  14.7 nm) and dND-PG-RGD-DOX (83.9  32.3 nm) in water was larger than that of dND-PG (36.4  11.8 nm). STEM image of dND-PG (Fig. 1B) showed that each dND-PG core consists of either single dND particle or a small cluster including a few primary particles of dND. In the STEM images of dND-PG-DOX and dND-PG-RGD-DOX (Fig. 1C, D), single dND particles were hardly found and the size of the clusters increased compared with that of dND-PG, implying that the hydrophobic DOX conjugated on the outmost surface favor interparticular interactions leading to weak aggregation. The content of DOX in dND-PG-DOX and dND-PG-RGD-DOX was quantified to be 5.6 wt% and 12.2 wt%, respectively, by measuring their absorbance at 480 nm. The drug release of dND-PG-DOX was studied at 37  C using pH 5.0 phosphate-citrate buffer and pH 7.4 PBS as media. As shown in Fig. 3D, 24.7% of DOX were released by 12 h when dND-PG-DOX was incubated at pH 7.4. Then, the released DOX increased slightly even after incubation for 48 h. However, when the pH was decreased to 5.0, the released DOX by 48 h was calculated to be 63.3%, which is 2.5-fold higher than the value under pH 7.4. This result indicates the preferential cleavage of hydrazone linkage

Fig. 6. Cell uptake of DOX, dND-PG-DOX and dND-PG-RGD-DOX following 24-hour treatment. Concentrations were normalized to DOX up to 4 mg/mL. Red fluorescence intensity was acquired on arithmetic scale and arithmetic means were used for quantification. Values are means  SD (n ¼ 6). Refer to footnote 1 in the text for explanation of linear range for DOX.

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

5399

Fig. 7. Cytotoxity of DOX, dND-PG-DOX and dND-PG-RGD-DOX assayed by CCK-8 reduction following 24-hour and 48-hour treatment. Concentrations were normalized to DOX up to 4 mg/mL. Values are means  SD (n ¼ 6).

under acidic condition [36,37]. From this point of view, dND-PGDOX is expected to show enhanced drug release in the lysosomes of cells (pH ¼ 5.0e5.5) but minimal premature drug release during circulation in bloodstream (pH ¼ 7.4). In in vitro cell experiments, the cytocompatibility of dND-PG and dND-PG-RGD was in the first place assayed. As shown in Fig. 4, neither A549 cells nor U937 macrophages showed any significant change in cell viability after 24-hour treatment of dND-PG or dNDPG-RGD at concentrations up to 400 mg/mL. Good cytocompatibility of both materials is thus suggested. Cell uptake of dND-PG and dND-PG-RGD was next evaluated. Uptake of the two materials in A549 cells and U937 macrophages after 24-hour treatment was quantified using FACS SSC intensity, an indicator proportional to cell granularity or internal complexity (Fig. 5 and Fig. S2). dND-PG showed little uptake in both A549 cells and U937 macrophages at most concentrations. A statistically significant but slight increase of dND-PG uptake only occurred at 200 mg/mL in A549 cells. The results suggest the capability of dNDPG against non-specific uptake by both non-phagocytic (A549) and phagocytic (U937) cells. We assumed that suppression of nonspecific cell uptake by the PG coating would provide a good basis for realization of highly selective drug delivery through a specific cell uptake mechanism via a targeting moiety. Indeed, this concept was substantiated by the cell uptake data of dND-PG-RGD. As mentioned above, RGD peptide (the targeting moiety) specifically binds to integrin receptor avb3 that is over-expressed in multiple types of malignant tumor cells. dND-PG-RGD was dose-

dependently taken up by A549 cells (integrin receptor avb3positive) but still largely evaded uptake by U937 macrophages (integrin receptor avb3 negative), only showing a slight increased uptake at 200 mg/mL. Subsequently, cell uptake of dND-PG-DOX and dND-PG-RGDDOX were compared against free DOX in A549 cells and U937 macrophages. Intracellular DOX levels were quantified in each type of cells after 24-hour treatment of the three materials with concentrations normalized to content of DOX up to 4 mg/mL (Fig. 6 and Fig. S3). All three materials resulted in a concentration-dependent increase in DOX levels indicated by the intensified fluorescence in both A549 cells and U937 macrophages. Regardless of cell type, cellular DOX level resulted from dND-PG-DOX at each concentration was less than half of that caused by free DOX within the linear range for DOX.1 On the other hand, dND-PG-RGD-DOX at each concentration resulted in a markedly higher DOX level than dNDPG-DOX in A549 cells that express RGD peptide receptors. But in U937 macrophages devoid of RGD peptide receptors, this material at each concentration only caused a slightly higher DOX level than ND-PG-DOX, which was still pronouncedly lower than that caused

1 Once DOX concentration reaches beyond the linear range for DOX, the fluorescence intensity will surpass the upper detection limit of the flow cytometer and the registered signals will not correlate linearly with DOX concentration any more but only approximate the upper detection limit (Fig. S3). Thus, the actual fluorescence signals for DOX should be much stronger.

5400

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

by free DOX. These data provide further concrete evidence for dNDPG’s effect against non-specific cell uptake as well as the targeting efficacy of RGD peptide. The mechanism of PG’s masking effect against cell uptake is still to be elucidated; there is study suggest that PG polymers act like PEG which can prevent protein adsorption [38]. On the basis of their cell uptake profiles, free DOX, dNDPG-DOX and dND-PG-RGD-DOX were then compared for cytotoxicity after being applied to A549 cells and U937 macrophages for 24 h and 48 h (Fig. 7). DOX showed non-selective toxicity to both A549 cells and U937 macrophages, though U937 macrophages were clearly much more sensitive to DOX than A549 cells. In sharp contrast, dND-PG-DOX appeared almost non-toxic to both types of cells even after 48 h of treatment. dND-PG-RGD-DOX displayed significant toxicity in A549 cells but little toxicity in U937 macrophages. Noteworthy is that cytotoxic effects of DOX and dND-PGDOX were not fully exerted after 24 h of treatment. The cytotoxicity data of free DOX, dND-PG-DOX and dND-PGRGD-DOX and their respective cell uptake profile are in good consistency, which is further evidenced by microscopic observations (Figs. 8e10, Figs. S4eS6). Mixed cell cultures of A549 and U937 were respectively incubated with the three materials with normalized DOX concentration of 2 mg/mL and photographed. Both free DOX and dND-PG-RGD-DOX resulted in pronounced red fluorescence in both types of cells after 30 min of incubation (Fig. 8, Fig. S4). dND-PG-RGD-DOX-treated cells were apparently brighter than free DOX-treated cells, indicating faster uptake. The fluorescence from dND-PG-RGD-DOX was primarily distributed in the cytoplasm, while fluorescence from free DOX was observed both in the cytoplasm and nuclei. In contrast, cells treated with dND-PGDOX appeared dim with much weaker red fluorescence. As incubation time reached 24 h, free DOX displayed appreciable toxic

effect on A549 cells (reduced cell number) and, more distinctly, on U937 macrophages (reduced cell number plus abnormal cell morphology) whereas dND-PG-DOX and dND-PG-RGD-DOX appeared to cause little apparent change in both types of cells (Fig. 9 and Fig. S5). In A549 cells, red fluorescence from dND-PGDOX was much weaker than that from free DOX and dND-PGRGD-DOX. Fluorescence of free DOX was stronger in the nuclei than in cytoplasm, while fluorescence from dND-PG-DOX and dNDPG-RGD-DOX appeared to be distributed essentially in the cytoplasm. In U937 macrophages, red fluorescence intensity followed the order: dND-PG-DOX < dND-PG-RGD-DOX < free DOX. Fluorescence of free DOX appeared concentrated with great brightness while red fluorescence from dND-PG-DOX and dND-PG-RGD-DOX appeared to be mostly cytoplasmic. After 48 h, U937 macrophages seemed to be annihilated by DOX while A549 cells survived (Fig. 10 and Fig. S6). The mixed cell culture either treated with dNDPG-DOX or dND-PG-RGD-DOX appeared to have similar morphology to the control cells. In the remaining A549 cells, red fluorescence from free DOX was clearly seen in the cytoplasm. Fluorescence from dND-PG-DOX and dND-PG-RGD-DOX was still distributed essentially in the cytoplasm. In U937 macrophages, red fluorescence ascribed to dND-PG-DOX and dND-PG-RGD-DOX was also mostly cytoplasmic. Cellular red fluorescence intensity followed the order: dND-PG-DOX < dND-PG-RGD-DOX < free DOX in both types of cells. DOX is a chemotherapy drug widely administered in the clinic with its anticancer action involving multiple mechanisms. Major mechanisms are interference with and damage of DNA, inhibition of macromolecular biosynthesis, stimulation of ROS production, apoptosis induction and membrane injury [39,40]. Blood cancers including lymphoma, leukemia and multiple myeloma are among

Fig. 8. A549 cells and U937 macrophages in mixed culture following 30-min treatment of DOX (A), dND-PG-DOX (B) and dND-PG-RGD-DOX (C). Concentrations were normalized to 2 mg/mL of DOX.A1, B1 and C1: bright field images. A2, B2 and C2: Green fluorescence is from A549 cells stained with CFDA-SE and blue fluorescence is from the nuclei of both A549 and U937 cells. A3, B3 and C3: Red fluorescence is from internalized DOX, dND-PG-DOX and dND-PG-RGD-DOX.

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

5401

Fig. 9. A549 cells and U937 macrophages in mixed culture following 24-hour treatment of DOX (B), dND-PG-DOX (C) and dND-PG-RGD-DOX (D). Concentrations were normalized to 2 mg/mL of DOX. A: Control cells. A1, B1, C1 and D1: bright field images. A2, B2, C2 and D2: Green fluorescence is from A549 cells stained with CFDA-SE and blue fluorescence is from the nuclei of both A549 and U937 cells. A3, B3, C3 and D3: Red fluorescence is from internalized DOX, dND-PG-DOX and dND-PG-RGD-DOX.

the cancer types most sensitive to DOX. The U937 cells used in our work are of the myeloid lineage and thus very sensitive to the toxicity of DOX, as shown in Figs. 7 and 10. However, when DOX was loaded onto dND-PG or dND-PG-RGD, its U937 cell-killing effect was totally prevented (Figs. 7 and 10), primarily because of PG coating’s shielding effect against cell uptake. Another reason might be that internalized dND-PG-DOX and dND-PG-RGD-DOX are sequestered from nuclear DNA in the cytoplasm. Indeed, microscopic observation provided some evidence therefor (Figs. 9 and 10). Another point to note is that the U937 cells in our work were differentiated into macrophages and stopped proliferation while A549 cells continued growth. DOX toxicity manifested as cell killing in U937 macrophages but seemed to be largely growth inhibition in A549 cells (Fig. 10). Sub-cellular localization of free DOX, dND-PG-DOX and dNDPG-RGD-DOX in A549 cells was closely observed after the cells were incubated with these materials (concentrations normalized to 2 mg/mL of DOX) for 24 h (Fig. 11). In free DOX-treated cells (Fig. 11AeD), red fluorescence was distinctly seen in both the nuclei and cytoplasm. Punctuated red fluorescence in the cytoplasm generally followed the distribution of blue lysosomal fluorescence. In dND-PG-DOX-treated cells (Fig. 11EeH), red fluorescence was

rather dim and only present in the cytoplasm but not nuclei, matching the pattern of blue lysosomal fluorescence. In dND-PGRGD-DOX-treated cells (Fig. 11IeL), red fluorescence was also only present in the cytoplasm but with great brightness. The above observations revealed that the acidic lysosomal compartment is a major depot of internalized dND-PG-DOX, dND-PG-RGD-DOX and DOX as well in A549 cancer cells. This is the rationale for using an acid-cleavable bond to connect DOX with dND-PG. Note that the fluorescence of dND-PG-RGD-DOX overlapped in part with the blue lysosomal fluorescence, indicating that DOX associated with dNDPG-RGD could dissociate and diffuse out of lysosomes (Fig. 11Ie K). Free DOX appeared to rapidly enter in the nucleus (Fig. 11AeD). The influence of the dND-PG/dND-PG-RGD platform on macrophage cell functions is another major subject of this study. Phagocytosis function, ROS generation and TNF-a secretion were analyzed in both resting and LPS-activated U937 macrophages after the cells were treated for 24 h with dND-PG and dND-PG-RGD at concentrations up to 200 mg/mL. In either resting or activated U937 macrophages, dND-PG only very slightly lowered the percentage of phagocytic cells (Fig. 12 and Fig. S7). ROS generation after dND-PG treatment showed little change in resting U937 macrophages but increased to a small extent in LPS-activated U937 macrophages

5402

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

Fig. 10. A549 cells and U937 macrophages in mixed culture following 48-hour treatment of DOX (B), dND-PG-DOX (C) and dND-PG-RGD-DOX (D). Concentrations were normalized to 2 mg/mL of DOX. A: Control cells. A1, B1, C1 and D1: bright field images. A2, B2, C2 and D2: Green fluorescence is from A549 cells stained with CFDA-SE and blue fluorescence is from the nuclei of both A549 and U937 cells. A3, B3, C3 and D3: Red fluorescence is from internalized DOX, dND-PG-DOX and dND-PG-RGD-DOX.

(Fig. 13 and Fig. S8). TNF-a secretion after dND-PG treatment was elevated to a small extent in resting U937 macrophages but was little affected in LPS-activated U937 macrophages. These results suggest that dND-PG has little influence on macrophage cell functions. The weak alteration in ROS generation and TNF-a secretion induced by dND-PG might be a non-specific macrophage reaction to phagocytosed particles. Similar to dND-PG, dND-PG-RGD only slightly lowered the percentage of phagocytic cells (Fig. 12 and Fig. S7) in both resting and LPS-activated U937 macrophages. ROS generation after dND-PG-RGD treatment was also unremarkable, showing a slight decrease in resting U937 macrophages and a slight increase in LPS-activated U937 macrophages (Fig. 13 and Fig. S8). Notably, TNF-a secretion after dND-PG-RGD treatment displayed a sharp increase in resting U937 macrophages but showed a tendency to decline in LPS-activated U937 macrophages, though not statistically significant (Fig. 14). The underlying mechanisms for this phenomenon are not clear but might be linked to the weak immunogenicity of RGD peptide. Cyclic RGD peptide is a wellestablished cell-binding motif for targeting malignant tumor cells over-expressing integrin-avb3 and has been applied to many nano devices for diagnostic imaging, drug delivery, gene therapy and etc. [41]. It has been reported that antagonism of integrin-avb3 by RGD

peptides is capable of inducing cell apoptosis at high concentrations [42]. However, we did not observe any remarkable cytotoxicity of dND-PG-RGD in our work (Fig. 4). On the other hand, it is thought that the exposed DOX on the nanoparticles could decrease the surface hydrophilicity and therefore negatively affect the cell uptake-evading effect of PG coating to some extent. Indeed, both dND-PG-DOX and dND-PGRGD-DOX resulted in a concentration-dependent increase in DOX levels in U937 macrophages (Fig. 6), which is not in keeping with the cell uptake profile of dND-PG and dND-PG-RGD (Fig. 5). A better drug loading strategy is being explored to improve this aspect. In addition, on the basis of results from the current work, further investigation is underway to characterize the in vivo pharmacokinetics of the dND-PG/dND-PG-RGD platform. 4. Conclusions We have demonstrated that dND-PG is a drug carrier platform with good cytocompatibility and a property against non-specific cell uptake, particularly by macrophages. When dND-PG is loaded with an anticancer drug (e.g. DOX) and led by an efficient targeting moiety (e.g. RGD peptide), it can realize highly preferential toxicity

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

to the intended tumor cells through specific uptake mechanisms with minimum uptake and toxicity in macrophages that are otherwise highly sensitive to the drug. dND-PG is advantageous for enhancing the efficacy of anticancer chemotherapeutics and reducing their toxicity by suppression of MPS elimination. Besides,

5403

dND-PG also provides a versatile platform amenable to various functionalizations for multiple theranostic purposes. From a fundamental perspective, the dND-PG/dND-PG-RGD platform described here proves the notion that selective cell uptake is best achieved on the basis of inhibition of non-specific uptake.

Fig. 11. Subcellular localization of internalized DOX (AeD), dND-PG-DOX (EeH) and dND-PG-RGD-DOX (IeL) in A549 cells. Blue fluorescence is from lysosomal staining by LysoTrackerÒ Blue. Red fluorescence is from internalized DOX, dND-PG-DOX and dND-PG-RGD-DOX. A, E, and I are bright field images. D: B merged with C. H: F merged with G. L: J merged with K.

5404

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

Fig. 11. (continued).

Fig. 12. Effects of dND-PG and dND-PG-RGD on the phagocytosis function of resting (A) and LPS-activated (B) U937 macrophages. Duration of treatment was 24 h. Percentage of phagocytic cells is calculated. Values are means  SD (n ¼ 6). Student’s t-test was performed (* and #p < 0.05, ** and ##p < 0.01).

Fig. 13. Effects of dND-PG and dND-PG-RGD on ROS generation in resting (A) and LPSactivated (B) U937 macrophages. Duration of treatment was 24 h. Values are means  SD (n ¼ 6). Student’s t-test was performed (** and ##p < 0.01).

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

5405

Fig. 14. Effects of dND-PG and dND-PG-RGD on TNF-a secretion in resting (A) and LPS-activated (B) U937 macrophages. Duration of treatment was 24 h. Values are means  SD (n ¼ 6). Student’s t-test was performed (** and ##p < 0.01).

Acknowledgments This work was partially supported by Grants-in-Aid for Scientific Research (No. 2400210802) from Japan Society for the Promotion of Science (JSPS). X. Chen also thanks the JSPS for providing the Postdoctoral Fellowship (No. P12108).

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2014.03.041.

References [1] Davies LC, Jenkins SJ, Allen JE, Taylor PR. Tissue-resident macrophages. Nat Immunol 2013;14:986e95. [2] Moghimi SM, Hunter C. Capture of stealth nanoparticles by the body’s defences. Crit Rev Ther Drug Carrier Syst 2001;18:527e50. [3] Ernsting MJ, Murakami M, Roy A, Li S-D. Factors controlling the pharmacokinetics, biodistribution and intratumoral penetration of nanoparticles. J Control Release 2013;172:782e94. [4] Moon JJ, Huang B, Irvine DJ. Engineering nano- and microparticles to tune immunity. Adv Mater 2012;24:3724e46. [5] Andersen AJ, Wibroe PP, Moghimi SM. Perspectives on carbon nanotubemediated adverse immune effects. Adv Drug Deliv Rev 2012;64:1700e5. [6] Alexander-Bryant AA, Vanden Berg-Foels WS, Wen X. Bioengineering strategies for designing targeted cancer therapies. Adv Cancer Res 2013;118:1e59. [7] Koo H, Huh MS, Sun I-C, Yuk SH, Choi K, Kim K, et al. In vivo targeted delivery of nanoparticles for theranosis. Acc Chem Res 2011;44:1018e28. [8] Saha K, Kim ST, Yan B, Miranda OR, Alfonso FS, Shlosman D, et al. Surface functionality of nanoparticles determines cellular uptake mechanisms in mammalian cells. Small 2013;9:300e5. [9] Amoozgar Z, Yeo Y. Recent advances in stealth coating of nanoparticle drug delivery systems. Wiley Interdiscip Rev Nanomed Nanobiotechnol 2012;4: 219e33. [10] Dominska M, Dykxhoorn DM. Breaking down the barriers: siRNA delivery and endosome escape. J Cell Sci 2010;123:1183e9. [11] Ishida T, Wang X, Shimizu T, Nawata K, Kiwada H. PEGylated liposomes elicit an anti-PEG IgM response in a T cell-independent manner. J Control Release 2007;122:349e55. [12] Wang L, Neoh KG, Kang E-T, Shuter B. Multifunctional polyglycerol-grafted Fe3O4@SiO2 nanoparticles for targeting ovarian cancer cells. Biomaterials 2011;32:2166e73. [13] Zhao L, Takimoto T, Ito M, Kitagawa N, Kimura T, Komatsu N. Chromatographic separation of highlysoluble diamond nanoparticles prepared by polyglycerol grafting. Angew Chem Int Ed 2011;50:1388e92. [14] Hofmann AM, Wurm F, Hühn E, Nawroth T, Langguth P, Frey H. Hyperbranched polyglycerol-based lipids via oxyanionic polymerization: toward multifunctional stealth liposomes. Biomacromolecules 2010;11:568e74.

[15] Arsalani N, Fattahi H, Laurent S, Burtea C, Elst LV, Muller RN. Polyglycerolgrafted superparamagnetic iron oxide nanoparticles: highly efficient MRI contrast agent for liver and kidney imaging and potential scaffold for cellular and molecular imaging. Contrast Media Mol Imaging 2012;7:185e94. [16] Kainthan RK, Hester SR, Levin E, Devine DV, Brooks DE. In vitro biological evaluation of high molecular weight hyperbranched polyglycerols. Biomaterials 2007;28:4581e90. [17] Maruyama K, Okuizumi S, Ishida O, Yamauchi H, Kikuchi H, Iwatsuru M. Phosphatidyl polyglycerols prolong liposome circulation in vivo. Int J Pharm 1994;111:103e7. [18] Calderón M, Quadir MA, Sharma SK, Haag R. Dendritic polyglycerols for biomedical applications. Adv Mater 2010;22:190e218. [19] Sisson AL, Steinhilber D, Rossow T, Welker P, Licha K, Haag R. Biocompatible functionalized polyglycerol microgels with cell penetrating properties. Angew Chem Int Ed 2009;48:7540e5. [20] Mochalin VN, Shenderova O, Ho D, Gogotsi Y. The properties and applications of nanodiamonds. Nat Nanotechnol 2012;7:11e23. [21] Manus LM, Mastarone DJ, Waters EA, Zhang XQ, Schultz-Sikma EA, Macrenaris KW, et al. Gd(III)-nanodiamond conjugates for MRI contrast enhancement. Nano Lett 2010;10:484e9. [22] Hegyi A, Yablonovitch E. Molecular imaging by optically detected electron spin resonance of nitrogen-vacancies in nanodiamonds. Nano Lett 2013;13: 1173e8. [23] Zhang X-Q, Lam R, Xu X, Chow EK, Kim H-J, Ho D. Multimodal nanodiamond drug delivery carriers for selective targeting, imaging, and enhanced chemotherapeutic efficacy. Adv Mater 2011;23:4770e5. [24] Zhang X-Q, Chen M, Lam R, Xu X, Osawa E, Ho D. Polymer-functionalized nanodiamond platforms as vehicles for gene delivery. ACS Nano 2009;3: 2609e16. [25] Wang L, Neoh KG, Kang ET, Shuter B, Wang SC. Superparamagnetic hyperbranched polyglycerol-grafted Fe3O4 nanoparticles as a novel magnetic resonance imaging contrast agent: an in vitro assessment. Adv Funct Mater 2009;19:2615e22. [26] Chang Y, Lee H, Chen K, Chang C, Tsai D, Fu C, et al. Mass production and dynamic imaging of fluorescent nanodiamonds. Nat Nanotechnol 2008;3:284e8. [27] Faklaris O, Joshi V, Irinopoulou T, Tauc P, Sennour M, Girard H, et al. Photoluminescent diamond nanoparticles for cell labeling: study of the uptake mechanism in mammalian cells. ACS Nano 2009;3:3955. [28] Mohan N, Chen CS, Hsieh HH, Wu YC, Chang HC. In vivo imaging and toxicity assessments of fluorescent nanodiamonds in caenorhabditis elegans. Nano Lett 2010;10:3692e9. [29] Zhao L, Xu Y-H, Qin H, Abe S, Akasaka T, Chano T, et al. Platinum on nanodiamond: a promising prodrug conjugated with stealth polyglycerol, targeting peptide, and acid-responsive antitumor drug. Adv Funct Mater; 2014. http:// dx.doi.org/10.1002/adfm.201304298. [30] Xiong J-P, Stehle T, Zhang R, Joachimiak A, Frech M, Goodman SL, et al. Crystal structure of the extracellular segment of integrin aVb3 in complex with an Arg-Gly-Asp ligand. Science 2002;296:151e5. [31] Graf N, Bielenberg DR, Kolishetti N, Muus C, Banyard J, Farokhzad OC, et al. aVb3 integrin-targeted PLGA-PEG nanoparticles for enhanced anti-tumor efficacy of a Pt(IV) prodrug. ACS Nano 2012;6:4530e9. [32] Zhao L, Shiino A, Qin H, Kimura T, Komatsu N. Synthesis, characterization, and magnetic resonance evaluation of polyglycerol-functionalized detonation nanodiamond conjugated with gadolinium(III) complex. J Nanosci Nanotechnol; 2014. In press.

5406

L. Zhao et al. / Biomaterials 35 (2014) 5393e5406

[33] Zhao L, Chano T, Morikawa S, Saito Y, Shiino A, Shimizu S, et al. Hyperbranched polyglycerol-grafted superparamagnetic iron oxide nanoparticles: synthesis, characterization, functionalization, size separation, magnetic properties, and biological applications. Adv Funct Mater 2012;22:5107e17. [34] Sintiprungrat K, Singhto N, Sinchaikul S, Chen S-T, Thongboonkerd V. Alterations in cellular proteome and secretome upon differentiation from monocyte to macrophage by treatment with phorbol myristate acetate: Insights into biological processes. J Proteomics 2010;73:602e18. [35] Quah BJC, Parish CR. The use of carboxyfluorescein diacetate succinimidyl ester (CFSE) to monitor lymphocyte proliferation. J Vis Exp; 2010:e2259. [36] Wang F, Wang Y-C, Dou S, Xiong M-H, Sun T-M, Wang J. Doxorubicin-tethered responsive gold nanoparticles facilitate intracellular drug delivery for overcoming multidrug resistance in cancer cells. ACS Nano 2011;5:3679e92. [37] Hu X, Liu S, Huang Y, Chen X, Jing X. Biodegradable block copolymerdoxorubicin conjugates via different linkages: preparation, characterization, and in vitro evaluation. Biomacromolecules 2010;11:2094e102.

[38] Siegers C, Biesalski M, Haag R. Self-assembled monolayers of dendritic polyglycerol derivatives on gold that resist the adsorption of proteins. Chem Eur J 2004;10:2831e8. [39] Fornari FA, Randolph JK, Yalowich JC, Ritke MK, Gewirtz DA. Interference by doxorubicin with DNA unwinding in MCF-7 breast tumor cells. Mol Pharmacol 1994;45:649e56. [40] Gewirtz D. A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin. Biochem Pharmacol 1999;57:727e41. [41] Delehanty JB, Boeneman K, Bradburne CE, Robertson K, Bongard JE, Medintz IL. Peptides for specific intracellular delivery and targeting of nanoparticles: implications for developing nanoparticle-mediated drug delivery. Ther Deliv 2010;1:411e33. [42] Chatterjee S, Matsumura A, Schradermeier J, Gillespie GY. Human malignant glioma therapy using anti-aVb3 integrin agents. J Neurooncol 2000;46: 135e44.