Delivery of Liposomal Quantum Dots via Monocytes for Imaging of

Feb 14, 2017 - *Phone: +972-2-6758658. Fax: +972-2-6757126. E-mail: ... However, a major concern is their biocompatibility, cytotoxicity, and fluoresc...
1 downloads 18 Views 9MB Size
Download PDF
Delivery of Liposomal Quantum Dots via Monocytes for Imaging of Inflamed Tissue Gil Aizik,† Nir Waiskopf,‡ Majd Agbaria,† Yael Levi-Kalisman,§,∥ Uri Banin,‡,∥ and Gershon Golomb*,†,∥ †

Institute for Drug Research, Faculty of Medicine, ‡Institute of Chemistry and the §Institute for Life Sciences, Faculty of Life Sciences, and ∥The Center for Nanoscience and Nanotechnology, The Hebrew University of Jerusalem, Jerusalem 9112001, Israel S Supporting Information *

ABSTRACT: Quantum dots (QDs), semiconductor nanocrystals, are fluorescent nanoparticles of growing interest as an imaging tool of a diseased tissue. However, a major concern is their biocompatibility, cytotoxicity, and fluorescence instability in biological milieu, impeding their use in biomedical applications, in general, and for inflammation imaging, in particular. In addition, for an efficient fluorescent signal at the desired tissue, and avoiding systemic biodistribution and possible toxicity, targeting is desired. We hypothesized that phagocytic cells of the innate immunity system (mainly circulating monocytes) can be exploited as transporters of specially designed liposomes containing QDs to the inflamed tissue. We developed a liposomal delivery system of QDs (LipQDs) characterized with high encapsulation yield, enhanced optical properties including farred emission wavelength and fluorescent stability, high quantum yield, and protracted fluorescent decay lifetime. Treatment with LipQDs, rather than free QDs, exhibited high accumulation and retention following intravenous administration in carotid-injured rats (an inflammatory model). QD−monocyte colocalization was detected in the inflamed arterial segment only following treatment with LipQDs. No cytotoxicity was observed following LipQD treatment in cell cultures, and changes in liver enzymes and gross histopathological changes were not detected in mice and rats, respectively. Our results suggest that the LipQD formulation could be a promising strategy for imaging inflammation. KEYWORDS: quantum dots, liposomes, monocytes, inflammation, carotid injury, nanomedicine, fluorescence imaging uantum dots (QDs) are nanometer-sized fluorescent semiconductor crystals, with a narrow emission band tunable by size through quantum confinement. QDs emerged in the past decade as superior substitutes for organic dyes in diagnostics due to their remarkable photochemical/ physical properties, such as high photostability, high quantum yield (QY), and remarkable capabilities for multiplexing.1,2 These properties, combined with methods to solubilize QDs in aqueous media as well as conjugating biological molecules, have led to a great interest for the utilization of QDs as fluorescent markers in molecular, cellular, and in vivo imaging.1,3 The toxicity of hydrophobic QDs due to organic ligands can be overcome by capping the QDs with a hydrophilic corona.4 Nevertheless, there are several unmet challenges in the development of QDs for theranostic use including (i) fluorescent signal reduction due to serum protein adsorption following systemic administration,5−7 (ii) enhanced quenching in the presence of oxidative or acidic conditions characterizing inflammation and lysosomal milieu,7,8 (iii) quenching due to intracellular aggregation,9 (iv) the intrinsic cytotoxicity10 of released Cd2+/Se2−, and (v) the wide systemic biodistribution. In an attempt to overcome some of the aforementioned limitations, nanocarrier systems for QD delivery have been

suggested including polymeric nanoparticles (NPs),11 micelles,12,13 and liposomes.14−21 However, none of these described delivery systems adequately addressed the requirements for efficient fluorescent imaging including high QY, prolonged decay lifetime, high stability in serum proteins and acidic pH, no premature leakage of QDs, high encapsulation yield, and demonstrating imaging efficacy in vitro as well as in vivo. Moreover, targeting QDs preferentially to the target tissue is required for specific and enhanced imaging and, at the same time, avoids off-target labeling as well as side effects. Inflammation plays a pivotal role in numerous human pathologies, including cardiovascular disorders,22 autoimmune diseases,23 and cancer.24 The inflammatory cascade is characterized by activation of the innate immunity system with substantial recruitment of neutrophils and monocytes to the diseased tissue.25 Inhibition of monocytes following preferred phagocytosis of charged particles has been shown by us and others as an effective therapeutic means in several

Q

© 2017 American Chemical Society

Received: January 2, 2017 Accepted: February 14, 2017 Published: February 14, 2017 3038

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

www.acsnano.org

Article

ACS Nano

Figure 1. Optical and structural characterizations of free and liposomal QDs in aqueous media. (a) Absorption and emission spectra of free QDs (CdSe/CdZnS) and LipQDs. Both free and LipQDs exhibited a narrow and symmetrical emission spectrum (peak at 655 nm). (b) Micrographs of 200 nM free or liposomal QDs viewed under visible or UV light. Representative TEM images of free QDs (c), empty positively charged liposomes (d), and LipQDs (e). Both LipQDs and empty liposomes demonstrate spherical shapes with narrow size distribution.

Table 1. Physicochemical Properties of LipQDs Examined in Vivo in Comparison to Free QDs and Empty Liposomes (Mean ± SD) mean hydrodynamic diameter (nm)

PDI

ζ-potential (mV)

13.4 ± 0.6 125.4 ± 0.4

0.56 ± 0.4 0.19 ± 0.01

23.7 ± 0.8

114.0 ± 0.3

0.25 ± 0.01

24.8 ± 0.5

free QDs empty liposomes LipQDs a

QD concn (nM)

lipid concn (mg/mL)

200

quantum yield (au)

fluorescent decay (ns)

0.58 ± 0.01

43

0.57 ± 0.01

35

17.4 ± 1.5 200

15a

Calculated value.

with high encapsulation yield and enhanced fluorescence stability for imaging. Liposomal QD Preparation. Encapsulation of the cargo in the aqueous core of liposomes is preferred for increased drug load. In addition, hydrophilic QDs are required for avoiding aggregation and preserving the spectral properties of the QDs in aqueous media. Hydrophilic, negatively charged CdSe/ CdZnS QDs were obtained by a previously described synthesis32 followed by replacement of the trioctylphosphine oxide (TOPO) hydrophobic coating of the QDs with glutathione using the ligand exchange method.33 This resulted in negatively charged QDs in physiological conditions because of the two carboxylic acid residues on every glutathione. Optical characterization of the obtained QDs showed preserved favorable characteristics of a broad excitation range, both narrow and symmetrical far-red emission spectra with a peak at 655 nm (Figure 1a,b), high QY (58%; Table 1),34 and a prolonged fluorescent decay lifetime (Table 1). The QDs exhibited a narrow size distribution of 13.4 ± 0.6 nm, as evidenced from the dynamic light scattering measurements (Table 1), and can be seen in the transmission electron microscopy (TEM) pictures (Figure 1c). Thus, the achieved

inflammatory-related disorders.26−28 We hypothesized that efficient fluorescence imaging could be achieved by a liposomal delivery system of DQs. We further hypothesized that monocytes, as part of the activated innate immunity system in inflammation,29 could be exploited as a courier of the liposomal QDs to an inflamed tissue. In this study, we have developed an efficient liposomal delivery system of QDs (LipQDs) for possible theranostic use in inflammatory-associated disorders. To the best of our knowledge, there are no reports on “biologic targeting” (i.e., via monocytes) of QDs. We describe here the formulation of QDladen targeted liposomes and the characterization of their physicochemical as well as stability and spectral properties. Imaging of inflamed tissue was examined in carotid-injured rats, an established model of restenosis.22 Restenosis, renarrowing of the artery following angioplasty, is characterized with initial recruitment of neutrophils followed by tissue infiltration and massive accumulation of monocytes into the inflamed arterial segment.30,31

RESULTS AND DISCUSSION In this study, we describe a liposomal QD delivery system, targeted to the inflammatory site via monocytes, characterized 3039

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

ACS Nano

Figure 2. Interaction between positively charged liposomes and QDs. (a) Schematic illustration depicting the interaction between negatively charged QDs and positively charged liposomes. (b) Gel electrophoresis of positively and negatively charged liposomes in agarose gel (dotted line indicates the bottom of the wells). The positively charged empty liposomes were spiked with different QD/DOTAP molar ratio (1:3.5k− 70k, respectively), and the negatively charged liposomes were spiked with QDs at a molar ratio of 1:3.5k QD/DSPG, respectively. In contrast to the migration of free QDs and QDs mixed with negatively charged liposomes to the positive pole (lanes 1 and 7, respectively), positively charged liposomes spiked with QDs stayed at the loading front (lanes 2−6). LipQDs showed the same migration pattern as spiked positively charged liposomes. Selected cryo-TEM images of empty liposomes spiked with QDs at different molar ratios of QD/DOTAP are shown in (c). A spherical morphology of vesicles was obtained with complete association between QDs and the vesicle’s membrane.

increasing the ionic strength of the dispersion by adding NaCl (Figure S1), which resulted in disassociation of the two oppositely charged molecules. An advantageous, very high encapsulation yield was achieved (80−90%), which is important for developing an efficient delivery system for theranostic use. Moreover, the relatively high QY and prolonged fluorescent decay lifetime of the QD particles were maintained in the liposomal formulation (Table 1). It should be noted that there were no free QDs (not encapsulated/associated with the vesicles) in the LipQD formulation, due to the strong electrostatic association between the QDs and the vesicles. This was verified by evaluating the properties of negatively charged liposomes. In contrast to empty negatively charged liposomes spiked with QDs, no free QDs were detected in size exclusion chromatography36 of spiked positively charged empty liposomes (data not shown). In addition, gel electrophoresis studies demonstrated no migration of free QDs to the positive pole when they were mixed with positively charged liposomes in a wide range molar ratio of QDs/DOTAP (Figure 2b). A similar migration pattern was observed with LipQDs (formulated with DOTAP). Only free QDs and QDs mixed with negatively charged empty liposomes migrated to the positive pole, indicating no

spectral properties make the hydrophilic QDs a suitable candidate for tissue-sensing applications. We took advantage of the negatively charged QDs to incorporate them into positively charged liposomes. The formulation obtained was composed of 1,2-distearoyl-snglycero-3-phosphocholine (DSPC)/cholesterol/1,2-dioleoyl-3trimethylammonium propane (DOTAP), with a molar ratio of 3:2:1, respectively. Consequently, liposomes with a homogeneous population of unilamellar, spherical-shaped vesicles, characterized by a size of 114 ± 0.3 nm and a low polydispersity index (PDI) of 0.25 ± 0.1, were obtained (Figure 1e and Table 1). The size of the liposomes enables effective phagocytosis35 of a filter-sterilized formulation. In addition, the liposomal QD formulation demonstrated an emission spectrum similar to that of free QDs (Figure 1a,b). To determine the encapsulation efficiency of QDs in the liposomal formulations spectrophotometrically, we had to overcome two problems: (i) light scattering of vesicles >100 nm and (ii) spectrophotometric overestimation due to the electrostatic interaction between the negatively charged QDs and the positively charged phospholipid, DOTAP. These obstacles were overcome by adding a detergent, octyl-β-Dglucopyranoside (OGP), which disrupted the vesicles, and by 3040

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

ACS Nano

Figure 3. Stability of LipQDs determined by structural and optical parameters in serum proteins and acidic pH. (a) Gel electrophoresis of free and liposomal QDs incubated with serum proteins or PBS for 24 h at 37 °C (dotted line indicates the bottom of the wells). In contrast to free QDs, LipQDs incubated in serum did not migrate to the positive electric pole, indicating stable association of the vesicles with the QDs with no QD leakage or disassociation. The optical stability of the LipQDs was further determined by quantum yield (b) and fluorescence decay (c,d) following 24 h incubation with serum proteins or acidic pH of 4.5 at 37 °C. The values in the bar chart (b) represent the percentage of fluorescence intensity normalized to free QDs in PBS (100%). LipQDs exhibited higher fluorescent stability in both milieus in comparison to free QDs (***P < 0.001). In contrast to free QDs with a t1/e of 8−9 ns, liposomal QDs preserved their prolonged fluorescence decay following incubation either in serum or at pH 4.5 (t1/e of 12 and 21 ns, respectively). The initial fluorescence intensity was normalized to 1, and t1/e is the time required to reach 1/e of the initial fluorescent signal.

our formulation, based on a strong electrostatic attraction between QDs and the liposome membrane, resulted in a stable complex of enhanced optical properties including a deep red emission wavelength, high QY, and protracted fluorescent decay lifetime (t1/e). The developed LipQD formulation characteristics represent a significant achievement in view of the undesirable alteration of QD optical properties previously reported.19 The inclusive and stable association of QDs with the lipidoic vesicles suggests a facile method for preparing liposomal QDs. This can be achieved by adding QDs to premade positively charged empty liposomes. This is because virtually all QD particles are associated with the vesicles due to the strong electrostatic attraction between the opposite charges of the QDs and the positively charged phospholipids. Structural and Fluorescence Stability of LipQDs. Aiming for an efficient diagnostic tool, it is obligatory to maintain high formulation stability in vivo, both structural and optical. The first concern in the trajectory of the LipQDs following systemic administration is the possible detrimental effect of serum proteins. Circulating proteins are known for their ability to interact with particulate delivery systems yielding a protein corona.6 Adsorption and/or interaction with serum proteins could impede the structural properties of the liposomes, resulting in premature leakage of the cargo and altered biodistribution.20,38 Incubation of LipQDs with serum proteins resulted in a significant increase of both size (203 ± 6 vs 114 ± 0.3 nm) and PDI value (0.76 ± 0.1 vs 0.25 ± 0.01;

electrostatic interaction between them (Figure 2b). Finally, even at highly competitive binding concentration of heparin (2500 IU/mL), which is several orders of magnitude higher than in vivo levels, only a small fraction of QDs disassociated from the vesicles (Figure S2a). This finding further confirms that the developed LipQD delivery system is characterized by a strong affinity between the QDs and the positively charged lipidoic vesicles. The inclusive association between QDs and the liposome membrane is further supported by the cryo-TEM images (Figure 2c). The QDs were shown to be in a close association with the vesicles’ membrane, with a negligible number of QDs not associated with the lipidoic vesicles. These results substantiate our working hypothesis that the strong electrostatic interaction between the opposite charges of the QDs (glutathione capping) and the lipid, DOTAP, enables the development of an efficient liposomal formulation. It should be noted that previous attempts to encapsulate hydrophilic QDs in the aqueous core of liposomes resulted in a very low payload.16 In an attempt to incorporate hydrophobic QDs within the lipid bilayer, only QDs with a size of 4−5 nm could be used (the thickness of the liposome bilayer membrane is ∼4 nm), thus restricting the emission wavelength to up to ∼500 nm (green spectrum).37 It is advantageous to use QDs with a near-infrared (NIR) emission spectrum (>650 nm) in order to overcome the tissue autofluorescence for efficient imaging applications in vivo. An additional disadvantage of former QD formulations is the premature leakage of QDs in vitro and/or in vivo. In contrast, 3041

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

ACS Nano

Figure 4. Effect of time and concentration on the internalization of liposomal QDs by monocytes in cell culture (RAW 264.7). Qualitative (a) and quantitative (b,c) assessment of cellular uptake by means of CLSM and FACS analyses, respectively. Free or liposomal QDs are shown in red, and cell nuclei are shown in cyan. The fluorescent intensity is normalized to nontreated cells; magnification is 60×. Note the constant degree of internalization, at each time point, exhibited by LipQDs (b, red lines), in contrast to free QDs that exhibited a dose−response pattern of cellular uptake (b, blue lines). Cellular uptake of liposomal QDs with a constant number concentration of vesicles (different number of QDs per vesicle) revealed a dose-dependent cellular uptake. Results are presented as the mean ± SD.

0.01, respectively; Figure 3b). It should be noted that in the acidic pH of the lysosome, LipQDs showed no sign of structural destabilization in terms of vesicle size and complex stability (Table S1 and Figure S2b). Similar to the results obtained upon incubation with serum proteins, a sufficient extended fluorescent decay lifetime was preserved in acidic conditions (Figure 3d). It can be summarized that the developed LipQD formulation preserves both high QY and extended t1/e in the presence of serum proteins and in acidic milieu of the lysosome. The efficient spectral protection endowed by liposomes, in both quenching conditions of acidic milieu and protein adsorption, is probably due to the decreased QDs’ surface area available for interaction. In addition to the known detrimental effects of acidic environment and protein adsorption, it has been shown recently that cations in physiological milieu, Ca2+, in particular, reduce the fluorescent intensity of QDs.41 We have examined the effect of Ca2+ cations (0.6 and 2 mM CaCl2, lysosomal and extracellular Ca2+ concentration, respectively) on the fluorescence intensity of free QDs in comparison to LipQDs (Figure S3). Incubation with Ca2+ ions only affected the fluorescence intensity of free QDs in solution (12% reduction with 2 mM, p < 0.001), whereas the fluorescent intensity of LipQDs was unaffected in both Ca2+ concentrations. The increased fluorescent stability could be explained by the shielding effect conferred by the lipidic membrane from the detrimental effect of positively charged ions because the liposome’s lipophilic bilayer is practically impermeable to ions. In addition, the cations are repelled by the positively charged membrane of the vesicles. Cellular Uptake, Spectral Stability, and Cytotoxicity of LipQDs. The prevalent approaches for achieving a high

Table S1). Nevertheless, no disassociation of the encapsulated QDs was detected, as evidenced by the absence of a free QD migration band in the electrophoresis experiment (after 24 h; Figure 3a). As expected, a shorter migration trail of proteinadsorbed QDs in comparison to QDs stored in PBS was observed. The reduced charge of LipQDs due to serum protein adsorption is supported by its band shape (Figure 3a, lane 3). The convex-shaped band of LipQDs, indicating attraction to the negative pole, was negated by adding serum proteins (Figure 3a, lane 4). It should be highlighted that the adsorption of serum proteins to the positively charged LipQDs, which affected their size and charge but not their binding affinity to QDs, could actually enhance the fluorescent intensity at the site of inflammation. This is because larger vesicles are internalized more avidly by circulating monocytes.39,40 Another detrimental effect due to serum protein adsorption is fluorescence quenching.5,7 Indeed, protein-adsorbed free QDs demonstrated a significant QY reduction by 78% in comparison to QDs in PBS (0.13 ± 0.01 and 0.58 ± 0.01, respectively; Figure 3b). LipQDs, however, exhibited higher QY following serum incubation (0.29 ± 0.01). It is known that the extended fluorescent decay lifetime (t1/e) of QDs, an important factor for increased signal-to-noise ratio,1 is significantly decreased following serum protein adsorption.19 Nevertheless, although serum proteins decreased LipQD t1/e, it was still sufficient for a high signal-to-noise ratio (Figure 3c). We also addressed the concern of reduced fluorescence in the lysosome (pH 4.5), due to possible etching of the ZnS shell layer and/or colloidal destabilization of the QDs.7,8,19,34 In contrast to free QDs that exhibit significant reduction of fluorescence following incubation in acidic pH, a high QY was preserved in the liposomal formulation (0.26 ± 0.01 and 0.42 ± 3042

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

ACS Nano

formulation fate after being phagocytized by monocytes, colocalization with the lysosomes was assessed using CLSM (Figure 5a). Although presenting different internalization paths,

accumulation of a drug delivery system at the target site are based on passive or active extravasation (enhanced permeability and retention effect or ligand-mediated).42−46 These approaches rely on extended circulation time typically imparted by PEGylation, resulting in the evasion of the shielded hydrophilic particles from the mononuclear phagocytic system (MPS).45 Our approach is quite the opposite. The LipQD formulation was specifically designed with optimal charge and size in order to achieve high uptake by circulating monocytes. Time- and dose-dependent uptake of LipQDs by RAW 264.7 cells, a murine monocyte/macrophage cell line, was assessed by confocal laser scanning microscopy (CLSM) and fluorescence active cell sorting (FACS) analyses (Figure 4a,b, respectively). Virtually all cells participated, to a certain extent, in phagocytizing either the free or liposomal QDs, after 1 h of incubation at all concentrations examined (Figure S4). In addition, free QDs, at all time points and concentrations examined, were internalized to a better extent than LipQDs. A time-dependent uptake was observed for both free and liposomal QD treatments (Figure 4b). However, dose-dependent internalization was observed only for free QDs (Figure 4b). This apparently suggests that the extent of LipQD phagocytosis is limited to a certain number of vesicles that can be internalized. In order to elucidate the mechanism of uptake, we treated the cells with different concentrations of QDs, while keeping the number concentration (Nc) of the vesicles constant (a constant number of vesicles but different concentration of QDs per vesicle, i.e., same number of vesicles at all different QD concentrations). This was done by adding different amounts of QDs (5, 10, and 20 nM) to a constant number of empty vesicles (15 mg/mL lipids). Indeed, a dose-dependent cellular uptake was revealed (Figure 4c). Altogether, taking the different experiments into consideration, the uptake mechanism of LipQDs, in contrast to that of free QDs, seems to be an active process (receptor-mediated) as expected from monocytes, which are professional phagocytic cells. On the other hand, the mechanism of free QD internalization is most probably macropinocytosis, receptor-independent,47 which characterizes the internalization of spherical particles under 100 nm. It is known that the physicochemical properties of a particulate system (e.g., shape, size, and surface charge) affect the internalization extent of particles.48,49 However, a direct comparison between free QDs (in fact, a colloidal solution) and LipQDs (a suspension) is improper. Thus, the higher extent of free QD accumulation in the cell culture could be explained by its conspicuous smaller size and by the different mechanism of cellular uptake. These findings suggest that in order to increase the cellular uptake of LipQDs, for achieving high accumulation at the inflamed region, the concentration of QDs in each vesicle should be increased rather than the number of vesicles. It should be noted that when escape from the MPS is desired, administration of a relatively high Nc of particles/vesicles is advised based on our observation. Empirical and calculation methods for determining the Nc of NPs has been described by us previously.50 The active phagocytosis of LipQDs supports our hypothesis of exploiting professional phagocytic cells of the innate immunity system, blood monocytes, as a courier for drugs or imaging agents. Regardless of the specific uptake mechanism, most particulate delivery systems are prone to degradation in the harsh environment of the lysosome, which is critical for the cargo functionality. In order to evaluate the liposomal

Figure 5. LipQD intracellular distribution and fluorescent stability in monocyte cell culture (RAW 264.7). (a) Qualitative assessment of liposomal QD (20 nM) colocalization with lysosomes following 5 h incubation. QDs are displayed in red, lysosomes are displayed in green (LysoTracker DND-26), and the colocalization is displayed in yellow. Note that almost all free or liposomal QDs are colocalized with lysosomal compartments. The quantitative determination of liposomal QD’s intracellular fluorescent stability is depicted in (b) using FACS analyses. Cells were treated for 5 h, washed with PBS and incubated with treatment-free medium for an additional 24, 48, and 72 h. LipQDs exhibited intracellular fluorescent stability higher than that of free QDs. The fluorescent intensities were normalized to the initial degree of uptake (immediately after the first 5 h of treatment) and to the number of cells. Results are presented as the mean ± SD, ***P < 0.001.

both free and liposomal QDs exhibited colocalization with the lysosomes 5 h after incubation. This observation is in accord with the known ultimate fate of internalized particles in lysosomal bodies, regardless of active or passive internalization mechanisms.47 Immediately after treatment of RAW 264.7 monocytes, a significant reduction of the fluorescence intensity was noted for both free and liposomal QDs (by 60−80 and 40−60%, respectively, Figures 5b and S5), which remained constant up to the 72 h time point. It was found that the liposomal formulation confers increased stability over time, in comparison to free QDs. These findings further substantiate the superior spectral stability of LipQDs over free QDs in different conditions including the presence of serum proteins, acidic conditions, and following cell internalization. It is clear that the cargo (LipQDs) should not be toxic to the courier (monocytes). The formulated LipQDs exhibited no cytotoxicity to monocytes following 48 h incubation (Figure S6). The massive emigration of circulating monocytes to the inflammatory site occurs within hours after insult.51 The noncytotoxic 3043

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

ACS Nano

Figure 6. Biodistribution of LipQDs in carotid-injured rats (a model of arterial inflammation). The phagocytosis of LipQDs by circulating monocytes 4 and 24 h after IV injection (200 nM, 2 mL) was determined by means of FACS analysis (a). Harvested blood monocytes were labeled with anti-CD68 antibody (FITC), and the percentage of QD-containing monocytes was assessed. Only LipQD treatment administered to both injured and intact animals resulted in marked internalization by circulating monocytes. The biodistribution of free or liposomal QDs in selected internal organs (liver, spleen, lungs, and kidneys) is shown in (b,c), analyzed by means of Typhoon scanner and ImageJ software, respectively. The fluorescent intensities were normalized to nontreated animals and are presented as the mean ± SEM, ***P < 0.001 (n = 4−6 animals in each group). Color scale bar: max = 14 815, min = 837.

effect observed after 48 h permitted in vivo studies for demonstrating effective accumulation of the engulfed QDs in monocytes for fluorescent detection of the inflammatory tissue. LipQDs in an Inflammatory Animal Model. The hypothesis of QD accumulation at the inflammatory site mediated by monocytes following endocytosis of liposomal QDs was examined in an inflammatory-associated model of vascular injury. The role of monocytes and their increased ingress in restenosis, renarrowing of the artery following angioplasty, has been documented in animals and humans.22,52−54 A significantly better uptake of LipQDs by monocytes was observed in both intact and injured animals in comparison to free QD-treated animals (Figure 6a). This finding validates our hypothesis that a particulate delivery system, such as liposomal QDs, will be internalized preferentially by circulating monocytes. The seeming discrepancy for the better uptake of free QDs in cell culture (Figure 4) is explained by the colloidal size of free QDs. Similar to dissolved molecules in the blood, which are unrecognized by professional phagocytic cells,53,55 very small particles such as QDs (∼15 nm) evade the MPS.56 It is well-known that reducing particles’ size to below 80 nm endows the injected particles to be “stealth” to circulating phagocytic cells, increasing their circulation residence time.56 The fundamental different pharmacokinetic profile of QDs and LipQDs is clearly manifested by their circulation time. Blood concentrations of free QDs remained constant up to 4 h after IV administration, whereas LipQDs presented a drastic concentration reduction a few minutes after administration (Figure S7a) as a result of effective phagocytosis by the MPS.

Increased accumulation of LipQDs, but not of free QDs, was observed in visceral organs rich with phagocytic cells, such as the liver, spleen, and lungs, 24 h after IV administration (Figure 6b,c). This is supported by the exclusive colocalization of QDs with tissue resident macrophages (Figure S8a,b). Only a weak and sporadic fluorescent signal was detected in the liver of animals treated with free QDs (Figure S8a). Interestingly, quantitative analysis of cadmium levels in rat’s visceral organs, following IV administration of free QDs, revealed that 60% of the injected dose was accumulated in the liver and spleen, similarly to LipQDs (Figure S7b). Thus, free QDs accumulated eventually in visceral organs but lost their fluorescence signal. The liver and spleen are known as the major disposal sites for particulate systems following systemic administration.56 It is suggested that free QDs accumulated in visceral organs due to coating with serum proteins (in accord with their adsorption capability; see Figure 3a). Moreover, it is plausible to suggest that protein-adsorbed QDs (as found above; Figure 3b) caused fluorescent quenching. These findings further highlight the superiority of LipQDs for imaging in vivo since free QDs are not phagocytized by monocytes/macrophages on one hand and are prone to quenching in the tissue on the other hand. Lastly, we evaluated the effectiveness of LipQD accumulation in the inflamed carotid artery model following IV injection. The time point for sacrificing the animals and examining the fluorescent signal of the arteries was based on the infiltration time of circulating monocytes post-injury (Figure S9). The number of resident phagocytic cells in intact arteries was negligible, whereas an increased number of phagocytic cells was observed post-injury, peaking at 24 h. This is in accord with previous findings that only macrophages originating from 3044

DOI: 10.1021/acsnano.7b00016 ACS Nano 2017, 11, 3038−3051

Article

ACS Nano

Figure 7. LipQD accumulation in the injured artery of rats 24 h post-treatment with LipQDs (200 nM, 2 mL). The injured carotid arteries (inset shows the surgical procedure) were harvested and scanned by means of a Typhoon scanner (a), and the fluorescent intensities were quantified by means of the ImageJ software (b). Fluorescent intensities were normalized to nontreated control and are presented as the mean ± SEM, ***P < 0.001 (n = 4−6 animals in each group). Color scale bar: max = 14815, min = 837. Note the high and selective QD accumulation in the injured artery only following LipQD treatment.

circulating monocytes are present in this model.57 QDs were not detected in the arteries of both intact and injured animal groups following free QD treatment as well as arteries of intact animal groups treated with LipQDs, as can be seen from the fluorescent imaging and immuno-histochemical examination (Figure 7a and Figure 8, respectively). In contrast, a significantly high fluorescent signal of LipQDs was observed in the injured arteries 24 h after IV administration (Figure 7a,b). Thus, a long retention time of 24 h, sufficient for image analysis procedures, was achieved. Our hypothesis that circulating monocytes, migrating to the inflamed tissue, can be exploited to deliver the phagocytized delivery system (“biologic” targeting) is supported by several direct and indirect evidence, (i) the liposomal formulation is charged, neither ultrasmall nor PEGylated; (ii) a short circulation time of LipQDs (