Gold Nanocluster-Mediated Cellular Death under Electromagnetic

Nov 8, 2017 - Anna Cifuentes-Rius†‡⊥⊗ , Angela Ivask†#⊗, Shreya Das†, Nuria Penya-Auladell†‡§, Laura Fabregas‡§, Nicholas L. Fle...
0 downloads 17 Views 6MB Size
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces 2017, 9, 41159-41167

Gold Nanocluster-Mediated Cellular Death under Electromagnetic Radiation Anna Cifuentes-Rius,†,‡,⊥,⊗ Angela Ivask,†,#,⊗ Shreya Das,† Nuria Penya-Auladell,†,‡,§ Laura Fabregas,‡,§ Nicholas L. Fletcher,∥ Zachary H. Houston,∥ Kristofer J. Thurecht,∥ and Nicolas H. Voelcker*,†,‡,⊥ †

ARC Centre of Excellence in Convergent Bio-Nano Science and Technology, Future Industries Institute, University of South Australia, Mawson Lakes Campus, Mawson Lakes Boulevard, Mawson Lakes, SA 5095, Australia ‡ Monash Institute of Pharmaceutical Sciences, Monash University, Parkville Campus, 381 Royal Parade, Parkville, VIC 3052, Australia § Grup d’Enginyeria de Materials (GEMAT), Institut Quimic de Sarria, Universitat Ramon Llull, Via Augusta 390, Barcelona 08022, Spain ∥ ARC Centre of Excellence in Convergent Bio-Nano Science and Technology, Centre for Advanced Imaging (CAI), The University of Queensland, Building 57, Research Road, St Lucia, QLD 4072, Australia S Supporting Information *

ABSTRACT: Gold nanoclusters (Au NCs) have become a promising nanomaterial for cancer therapy because of their biocompatibility and fluorescent properties. In this study, the effect of ultrasmall protein-stabilized 2 nm Au NCs on six types of mammalian cells (fibroblasts, B-lymphocytes, glioblastoma, neuroblastoma, and two types of prostate cancer cells) under electromagnetic radiation is investigated. Cellular association of Au NCs in vitro is concentration-dependent, and Au NCs have low intrinsic toxicity. However, when Au NC-incubated cells are exposed to a 1 GHz electromagnetic field (microwave radiation), cell viability significantly decreases, thus demonstrating that Au NCs exhibit specific microwave-dependent cytotoxicity, likely resulting from localized heating. Upon i.v. injection in mice, Au NCs are still present at 24 h post administration. Considering the specific microwave-dependent cytotoxicity and low intrinsic toxicity, our work suggests the potential of Au NCs as effective and safe nanomedicines for cancer therapy. KEYWORDS: gold nanoclusters, microwave frequency, electromagnetic radiation, hyperthermia, cytotoxicity Au NP heating use a RF generator called “Kanzius Machine” (Therm-Med LLC, Erie, PA), which operates at 13.56 MHz and at relatively high power output (up to 1000 W).6 Teams of chemists, radio-engineers, and academic surgeons have investigated the use of this machine and Au NPs 5−100 nm in diameter to kill cancer cells.8,11 They proved that Au NPs can convert electromagnetic energy into heat under a 13.56 MHz EM field, and the efficacy of heating depends on concentration and diameter of the Au NPs.9,11 The localized heat released from Au NPs under RF denatures proteins, disrupts lipid bilayers, and results in irreparable damage to intracellular structures and organelles of cancer cells.9 Other authors have used microwave frequency EM radiation to induce the heating of Au NPs. Hamad-Schifferli et al. found that hairpin-loop DNA covalently attached to 1.4 nm diameter gold nanocrystals was able to dehybridize into solution when exposed at 1 GHz EM field (microwave range) due to the localized heat generated.12 On the other hand, the heat release by Au NPs in the microwave range has been shown to be very

1. INTRODUCTION The possibility of using localized nanoparticle heating to remotely control biological functions has not gone unnoticed in the field of drug delivery, and it has been explored for several biomedical applications, including cancer therapy and diagnostics as well as triggered drug/gene release and treatment of bacterial infections.1−5 The thermal dissipation of nanosize gold (Au NPs) under electromagnetic (EM) radiation allows noninvasive, triggered, and localized therapies and can be induced by both infrared and radiofrequency/microwave radiation.6 Infrared light is commonly used for photothermal therapy, as infrared lasers and visible light can excite the surface plasmon resonance of Au NPs exploiting a well-understood mechanism.7 EM irradiation between the radiofrequency (RF) and microwave regions for hyperthermia in cancer treatment has gained special attention in the last 10 years due to its ability to penetrate deeper within nonconducting materials.8,9 However, the heating mechanism of Au NPs under RF and microwave irradiation is still not fully understood.6,10 In a recent review by Ackerson et al., two independent research lines based on different operating frequencies, RF or microwave, are categorized. The review revealed that most papers reporting © 2017 American Chemical Society

Received: August 30, 2017 Accepted: November 7, 2017 Published: November 8, 2017 41159

DOI: 10.1021/acsami.7b13100 ACS Appl. Mater. Interfaces 2017, 9, 41159−41167

Research Article

ACS Applied Materials & Interfaces

2. EXPERIMENTAL SECTION

localized, as recently proven by Kabb et al. In this work, they demonstrated, using a “polymeric thermometer”, that the heat generated by 5 nm Au NPs under microwave is localized within a radial distance from the nanoparticle surface less than 2 nm and is quickly dissipated at longer distances.13 This study by Kabb et al. was the first one to directly prove the surfacespecific heating from the nanoparticles, rather than the increase of temperature from the bulk solution. We have also recently indirectly demonstrated remote heating of thermoresponsive nanomaterials by microwave.14 The efficacy of microwave EM field in heating tissues is also already clinically proven. It has been shown that microwave frequencies (typically 915 MHz or 2.45 GHz) can heat tissues to cytotoxic level using an applicator antenna.15,16 It must be noted, however, that these studies with applicator antenna have been conducted without any addition of nanoparticles, and thus, the localization of heat has been controlled only by positioning the antenna. Addition of specifically localized nanoparticles to such a system would increase the precision of heating and instead of an invasive antenna would allow to use externally created microwave EM field. To our knowledge, microwave-induced heating of Au NPs for clinical cancer hyperthermia has not been explored.4,6 Among Au NPs, gold nanoclusters (Au NCs) have attracted great attention in the scientific community as they are atomically monodispersed subnanometer particles (655 nm, respectively. For ICP-MS, the cells were digested with aqua regia (3:1 concd HCl/HNO3) (900 μL of aqua regia was added to 100 μL of cells) overnight. Alternatively, Au NC exposed HR1K cells were chemically etched prior the digestion to remove cell surface bound nanoparticles. Cells were treated with 0.034 mM/0.2 mM I2/KI mixture for 1 min, washed, and digested as above. Digested samples were then diluted in a diluent containing 2.4% HCl, 0.04% HNO3, and 0.5% thiourea, and Au content was analyzed with a qqq-ICP-MS (Agilent 8800). Mass of Au per cell was calculated per cell; for that, the number of cells were counted before EM exposure. Adherent cells that were trypsinized from glass surface and suspension cells were counted using hemocytometer or flow cytometer. 2.5. Biodistribution of Au NCs in Vivo. All studies were in accordance with the guidelines of the Animal Ethics Committee of The University of Queensland (UQ; Approval AIBN/CAI/530/15/ ARC/NHMRC) and the Australian Code for the Care and Use of Animals for Scientific Purposes. Healthy male BALB/c nude mice (∼20 g; Animal Resources Centre, Australia) from 8 weeks old were imported into the CAI animal holding facility and monitored for 1 week prior to the study in order to acclimatize to the environment. All animals were provided with free access to food and water before and during the imaging experiments. Mice were anesthetized using 2% isoflurane (gradually reduced as required to maintain anesthesia and respiratory rate) in oxygen at a flow rate of 2 L/min, and Au NCs (0.8 mg/mL) were administered via tail vein i.v. injection (29G) in 200 μL of phosphate-buffered saline. Mice were then anesthetized (as above) and fluorescence images acquired at 0.5, 2, 4, 6, and 24 h postinjection using an in vivo MS FX Pro instrument (now supplied by Bruker Corp.) with mice in both prone and supine orientations. Au NC fluorescence images were collected with 480 ± 10 excitation filters, respectively, using a 700 nm ±17.5 nm emission filter set (f-stop 2.80, 4 × 4 pixel binning, 190 mm field of view, and 5 s exposure time). X-ray images (f-stop 2.80, 0.2 mm aluminum filter, 190 mm FOV, 10 s acquisition time) were acquired to provide anatomical registration. All images were processed using Image-J software (National Institutes of Health). Fluorescence images were false colored and overlaid onto X-ray images. 2.6. Ex Vivo Imaging of the Main Organs. At 24 h post injection, blood samples were collected via heart puncture (29G) before the mice were sacrificed via cervical dislocation. Major organs (heart, lungs, spleen, liver, kidneys, and gastrointestinal tract) and blood were imaged ex vivo using the same acquisition settings as above. Semiquantitative region of interest (ROI) analysis of ex vivo fluorescence images was conducted utilizing Bruker MI SE software to obtain values of sum fluorescence intensity per organ. 2.7. Quantification of Au Content Accumulated in each Organ. The heart, lungs, spleen, liver, and kidneys of each mouse was weighted and mechanically homogenized through a 70 μm cell strainer using cold 2% (w/v) bovine serum albumin and 1% (w/v) sodium azide in phosphate-buffered saline (PBA) to suspend the cells. Organ

cell suspensions and blood samples were centrifuged (500g for 5 min and 200g for 10 min, respectively), the supernatant was removed, and the remaining cell pellets were resuspended in cold 50:50 4% (w/v) paraformaldehyde/PBA and incubated in the dark at 4 °C overnight. Fixed cell suspensions were then centrifuged (500g for 5 min), the supernatant was removed, and each pellet was resuspended in 1 mL of cold PBA. Cell suspensions were then digested in aqua regia as described previously until full digestion of the organ cell suspensions was achieved. Au content in digested samples was analyzed by ICP-MS following the same protocol previously described. The amount of Au per organ (in mass) per each replicate and its average were calculated and plotted. 2.8. Exposure of Cells to Microwave Irradiation in Vitro. Prior to microwave irradiation, HR1K suspension cells and 3T3, U87, SHSY5Y, PC3, and C4−2B adherent cells were exposed to Au NCs as described above. A control without Au NCs was included for all cell lines (n = 3). Cells without Au NCs (n = 3) were also exposed to microwave in order to study the effect of the electromagnetism on cells without Au NCs. For microwave exposure, adherent cells on coverslips were added to sterile microcentrifuge tubes into which 500 μL of cell culture media was added (so that the glass slide was soaked in the cell medium). In the case of suspension cells, 500 μL of cells were pipetted to microcentrifuge tubes. Microcentrifuge tubes with cells were then exposed to 1 GHz (microwave range) electromagnetic radiation for 8 min for all cell lines except the C4−2B prostate cancer cell line that was exposed to the microwave field for 7 min (optimization experiments were carried out within a 3−15 min range). A microwave field was generated using a custom-made EM generator described elsewhere (Supporting Information, Figure S6).14 Between the successive irradiations, a 2 min interval was applied to cool the system. During the microwave irradiation process, the cells were allowed to stay in the closed microcentrifuge tubes not more than 60 min (no significant cell death was observed when either adherent or suspension cells were kept outside the temperature and CO 2 controlled environment for 60 min; data not shown). 2.9. Assessment of Cellular Viability after Microwave Irradiation. After microwave irradiation, the cells were placed back on the microplate at 37 °C, 5% CO2. The B-lymphocyte suspension cells were pipetted to 48-well microplates, and glass slides with adherent cells were placed onto a 24-well plate. In the case of adherent cells, cell culture medium that was used in microcentrifuge tubes at the time of microwave irradiation was placed over the slides with cells. The cells were incubated at 37 °C and 5% CO2 for 24 h. The viability of adherent cells was assessed by resazurin assay (resazurin reagent was added to cell culture medium at 30 μg/mL, cells were incubated for 4 h and fluorescence at 530 nm excitation/590 nm emission was measured). The viability of suspension cells was assessed by staining the cells with fluorescein diacetate (FDA, final concentration 0.5 μg/ mL) and propidium iodide (PI, final concentration 5 μg/mL) (cell staining was performed in PBS for 15 min at room temperature) and analyzing the cells with an imaging flow cytometer (Amnis Image Stream X). At least 10000 events were collected, populations of live cells (stained with FDA, 488/525 nm) and dead cells (stained with PI, 488/590 nm) were selected, and the percentage or each cell population was calculated. In parallel to microwave-irradiated cells, the viability of nonirradiated cells was also analyzed using analogous methods. 2.10. Detection of Apoptosis in Microwave-Irradiated Cells. An annexin V-FITC apoptosis detection kit (Sigma) was used as suggested by the manufacturer. HR1K suspension cells that had been exposed to 1−100 μg/mL AuNCs for 24 h and subsequently treated with microwave (or not treated) were incubated in Annexin V binding buffer (Sigma) with Annexin V FITC conjugate (100× dilution from Sigma kit) and propidium iodide (50× dilution from Sigma kit) for 10 min. The cells were analyzed using an Amnis imaging flow cytometer where apoptotic cells were identified using 488 nm excitation/480− 560 nm emission and late apoptotic plus necrotic cells using 488 nm excitation/595−660 nm emission. 41161

DOI: 10.1021/acsami.7b13100 ACS Appl. Mater. Interfaces 2017, 9, 41159−41167

Research Article

ACS Applied Materials & Interfaces

salt concentrations (up to 2.5 M) with a dH of 11.6 ± 0.5 nm (Figure 1D, blue line). The measurement remained the same when repeated at day 5 and 12 (Figure 1D, red and green lines, respectively), showing the colloidal stability of the BSA-coated Au NCs. The colloidal stability was also confirmed after measuring the zeta (ζ) potential, which resulted in 33−35 mV. Additionally, their stability in physiological conditions was also confirmed at 4 and 37 °C by performing the same experiment at these temperatures (Supporting Information, Figure S1). DLS measurement of Au NCs in PBS at room temperature gave a dH of 9.1 nm (Pdl of 0.264), similar to what was observed in the salt-aggregation study. Interestingly, the dH of Au NCs equals to the one observed for BSA alone.33 This observation matches with the studies that observed how BSA acts as a template for the formation and stabilization of Au NCs without significantly affecting its secondary structure.34−36 The BSA/Au ratio is key to achieve subnanometer-sized clusters because BSA acts as a cluster template, and the formation of Au NC is only possible when (a) enough BSA is in the solution and (b) the pH is raised above 10.17 During formation of Au NCs, first BSA coordinates Au3+ ions through the histidine groups within its structure and after rising the pH, the tyrosine residues are able to reduce the Au3+ ions in a controlled way, while the cysteine amino acids stabilize the formed cluster via disulfide groups. We confirmed the importance of the BSA concentration and the pH of the reaction by exploring the synthesis at different conditions (Figure S2, Supporting Information). We observed that, with the same amount of Au 3+ (10 mM), the minimum concentration of BSA in which the Au NCs were successfully formed was 50 mg/mL (Figure S2a), which was used throughout the study for the synthesis of Au NCs. We also proved that pH of the reaction less than 11 did not result in formation of high-quality Au NCs. At pH less than 8, no Au NCs were produced, whereas at pH 8, heterogeneous Au NCs with relatively low fluorescence intensity were formed (Figure S2b). 3.2. Association of Au NCs in Vitro with Different Cell Types. Due to the ability of Au NPs to induce hyperthermia only in very close vicinity, at some nanometers from the particles surface,13 cellular localization of those particles is crucial when their cytotoxicity under EM is studied. To qualitatively and quantitatively assess the amount of cellassociated Au NCs, either cell surface bound or internalized Au NCs, we used fluorescence imaging and inductively coupled plasma mass spectrometry (ICP-MS) (Figure 2). Cellular association of Au NCs was studied using mouse fibroblasts (3T3) and five cancerous cell lines from human: Blymphocytes from lymphoma (HR1K), U87 from glioblastoma, SHSY5Y from neuroblastoma, PC3 and C4−2B from prostate. Using confocal microscopy, we tracked Au NCs’ intrinsic fluorescence (488 nm excitation/∼650 nm emission) associated with cells. In Au NC-exposed cells, red fluorescing areas corresponding to Au NCs were visible in and around the cells (Figure 2A2,B2,C2,D2,E2,F2; these areas are also emphasized with white arrows on Figure 2A3,B3,C3,D3,E3,F3). ICP-MS results show that, in general, the cell-associated mass of Au was 170−225 fg per each cell that was exposed for 24 h to 100 μg of Au NCs/mL and 100−230 fg for those exposed to 50 μg of Au NCs/mL. Below 50 μg of Au NCs/mL exposure concentration, the cell-associated mass of Au decreased linearly (Figure 2A4,B4,C4,E4,F4). Although the amount of Au NCs associated with each cell correlated well with Au NC exposure

3. RESULTS AND DISCUSSION 3.1. Synthesis and Characterization of Fluorescent Au NCs. We synthesized Au NCs templated by BSA given its biocompatibility, stealth properties, and the possibility of carrying out further functionalization with targeting moieties.26,31 The synthesized Au NCs stabilized by BSA molecules as reported by Xie et al.26 showed an intense red fluorescence emission under UV light and an emission peak at around 660 nm when excited at 488 nm (Figure 1A and B). As expected, no

Figure 1. Characteristics of Au nanoclusters (Au NCs). (A) Appearance under visible (top images) and UV light (bottom images); Au NC concentration was 700 μg/mL, and water and BSA solutions were used for comparison. (B) Absorbance and fluorescence (after excitation at 488 nm) spectra of 100 μg/mL Au NCs. (C) TEM view of Au NCs (scale bar 20 nm); inset shows magnified view of the NCs (scale bar 1 nm). (D) Salt-induced aggregation of Au NCs over time measured by DLS.

surface plasmon resonance was observed at this particle size (Figure 1B). TEM showed that the formed Au NCs were 1.7 ± 0.4 nm in core diameter with a narrow size distribution (Figure 1C). The hydrodynamic diameter (dH) of Au NC was also measured by dynamic light scattering (DLS), and their colloidal stability in front of salt concentration and time was investigated (Figure 1D). Colloidal stability in physiological conditions is paramount to avoid agglomeration of nanoparticles, which can have a direct impact on cellular internalization complicating dosimetry.32 In particular, salt (NaCl), which is present at high concentation in the cell culture media, is one of the main reasons why particles agglomerate.30 Because Au NCs were suspended in osmotically and pH-controlled buffer (PBS 1× pH 7.4) before in vitro and in vivo studies, the effect of salt concentration present in the media was explored. Since we know that an increase of dH is likely due to nanoparticle aggregation, we studied the effect salt-induced aggregation of Au NCs using DLS. Therefore, the colloidal stability of Au NCs under physiological conditions can be proved.30 The DLS measurement was carried out immediately after adding the NaCl solution to the Au NC solution (in Milli-Q water), and again 5 and 12 days later, since aggregation can be a timedependent process. The dH was initially 8.8 ± 0.5 nm in the presence of NaCl 0.02 M which remained constant at higher 41162

DOI: 10.1021/acsami.7b13100 ACS Appl. Mater. Interfaces 2017, 9, 41159−41167

Research Article

ACS Applied Materials & Interfaces

Figure 2. Cellular association of Au NCs during 24 h exposure. Confocal fluorescence microscopy images of both nonexposed and Au NC exposed HR1K (A1-A3), U87 (B1−B3), SHSY5Y (C1−C3), 3T3 (D1−D3), PC3 (E1-E3), and C4−2B (F1−F3) cells (scale bar: 10 μm). In nonexposed cells (column 1; A1−F1) and Au NC exposed cells (column 2; A2−F2), nuclei are stained blue with Hoechst 33342, cytoskeleton yellow and Au NCs show red. Column 3 (A3−F3) is identical with column 2 but shows only nuclei and Au NCs. White arrows indicate red fluorescence originating from Au NCs. Column 4 shows the quantification of cell-associated Au with ICP-MS in HR1K (A4), U87 (B4), SHSY5Y (C4), 3T3 (D4), PC3 (D4), and C4−2B (F4) cells at different Au NC exposure concentrations (n = 3).

concentration-dependent increase in cellular uptake of Au NCs up to 6 μg Au NCs/mL after which the intracellular concentration of Au NCs remained constant and the percentage of cell-surface-bound Au NCs increased (Figure S3). For example, from cells that were pre-exposed to 100 μg/ mL of Au NCs, approximately 66% of Au initially associated with the cells was removed by chemical etching. In general, in HRIK cells 90−34% of the Au NCs, depending on the exposure concentrations, was internalized by the cells while the rest remained bound on cell surface. 3.3. In Vivo Biodistribution of Au NCs. Elucidation of the in vivo biodistribution of nanomaterials is a key preliminary study to understand their potency under physiological conditions and bioavailability. Since Au NCs exhibit NIR fluorescence, their in vivo biodistribution can be conveniently tracked by fluorescence imaging. In this study, we performed both in vivo and ex vivo fluorescence imaging of intravenously

concentration, our data indicate that, in general, Au NCs associated with cells at a relatively low efficiencyonly 0.1− 0.3% of the total mass of Au to which the cells were exposed during 24 h was found in association with cells in ICP-MS analysis. This may be due the high colloidal stability of Au NCs at 37 °C as demonstrated previously (Figure S1b). Because Au NCs did not sediment at the bottom of the well where the cells were seeded, the Au NCs−cell interaction was diminished, and therefore, a low amount of Au NCs was able to be internalized by the cells.32 On the other hand, even the percentage of cellassociated Au NCs appeared relatively low; 170−225 fg Au per cell corresponds to 20−30 × 106 Au NCs. In order to decipher the nature of cellular association with Au NCs, i.e., if the clusters were intracellular or externally bound, we applied to Au NC-exposed HR1K cells a chemicaletching procedure that according to Cho et al.37 causes removal of extracellularly bound Au particles. The results show a 41163

DOI: 10.1021/acsami.7b13100 ACS Appl. Mater. Interfaces 2017, 9, 41159−41167

Research Article

ACS Applied Materials & Interfaces

We suggest that Au NC fluorescence in those organs could not be seen ex vivo (Figure S4b) due to the low limit of detection of the methodology. Accumulation of Au NCs in liver and spleen may be due to the uptake of clusters through macrophages residing in the mononuclear phagocyte system (MPS) as has been observed for others nanomaterials.39 3.4. In Vitro Cytotoxicity of Au NCs under Microwave Irradiation. The final goal of this study was to assess the ability of Au NCs to induce cellular hyperthermia and cellular death under microwave irradiation. Before that, we measured the increase in bulk solution temperature relative to a control without Au NCs when exposed to microwave. Despite concentrating the Au NC solution to 2 mg/mL, the temperature of the control was only ∼2 °C below the measured for the Au NC solution. Kabb et al. directly measured the temperature of the nanoenvironment of microwave-induced heating of Au NP using a “polymeric thermometer”.13 This work found that the local temperature around the Au NP increased up to 70 °C, while the temperature of the bulk solution temperature only reached 40 °C. However, the temperature differential decreased drastically at a greater distance than 2 nm away from the particle surface. These results match our observations, suggesting that the heating happens locally within the Au NCs surface and then it is quickly dissipated to the bulk solution. We hypothesized then that Au NCs would specifically induce localized cytotoxicity in EM fields while being harmless otherwise and, hence, first evaluated the cellular effects of those particles in the absence of EM radiation. As expected,40,41 the viability of cells that had been exposed to varying concentrations of Au NCs (up to 100 μg/mL) was high (around 90%) (Figure 4A). To assess the cytotoxicity of Au NCs under microwave field, we first ensured that the EM radiation alone did not affect the cellular viability. In general, increasing microwave radiation times resulted in increased cellular death (Figure S5). For example, 10 min exposure to 15 W microwave radiation resulted in death of 50% HR1K Blymphocytes and in death of 32% U87 glioblastoma cells. We suggest that this unspecific cell death was due to the rise in temperature of the copper coil in the EM setup (described in Figure S6), which underwent heat exchange with the cell culture medium. Our optimization study (Figure S5a−c) showed that the largest difference between cell viability in nonexposed and Au NC-exposed cells under microwave radiation was achieved when the microwave radiation lasted for 8 min; this radiation time was selected for all further experiments except for C4−2B prostate cancer cell line which remained unaffected by the microwave radiation at a maximum time of 7 min (Figure S5d). Therefore, at the chosen microwave exposure times, and despite the unspecific heating of the cell culture media due to the heating of the coil, there was no significant cell death observed in cells without Au NCs. The cytotoxicity of Au NC-exposed B-lymphocyte (HR1K), glioblastoma (U87), neuroblastoma (SHSY5Y), fibroblast (3T3), and prostate cancer cells (PC3 and C4−2B) after 8 min of microwave radiation, and 7 min for C4−2B cells is shown in Figure 4B. We observed that under microwave radiation, all of the cells suffered from cellular death or inhibition, and this inhibitory effect was dependent on Au NC exposure concentration. For example, at the highest Au NC exposure concentration (100 μg/mL) where the cellular content of Au was ∼170−240 fg per cell (Figure 2), 40−60% of the cells were killed or inhibited (Figure 4). With decreasing

(i.v.) administered Au NCs to mice (8 μg/g) to semiquantitatively assess their biodistribution and bioavailability. Real-time mouse images were taken at different time-points after injection, similar to other biodistribution studies done for Au NCs and other nanomaterials.31,38 We observed that Au NCs were still present in the mouse 24 h post i.v. administration (Figure 3A) since there is fluorescence signal observed, suggesting good bioavailability.

Figure 3. Distribution of Au NCs in mice. Au NCs were administered i.v. through the tail vein. (A) In vivo fluorescence imaging of mice in supine (top panels) and prone (lower panels) position are shown; false color scale is indicative of Au NC concentration. (B) ICP-MS analysis of the organs of Au NC-injected mouse, performed after homogenization of the separated organs.

Because Au NCs did not show specific organ accumulation in vivo, ex vivo imaging was used to further probe biodistribution. We collected liver, kidneys, spleen, lungs, heart, the gastrointestinal tract, and a drop of blood and analyzed their fluorescence ex vivo (Figure S4a). In Au NC-exposed mice, higher fluorescence intensity was observed in the liver compared to the control. High fluorescence detected in the gastrointestinal tract of Au NCinjected mice was also present in control mice and, therefore, can be attributed to autofluorescence. A semiquantitative region of interest (ROI) analysis of the fluorescence signal of each organ of the three injected mice was performed (Figure S4b). For the other organs, the fluorescence levels in the Au NCinjected mice were similar to the organs harvested from control mice (slight but not statistically significant increase in fluorescence was observed in liver of Au NC-exposed mice). In order to quantify the amount of Au accumulated in each organ, the organs were homogenized and digested in aqua regia for ICP-MS analysis (Figure 3B). The results showed that Au NCs accumulated preferably in liver where the concentration of Au reached 12164 ± 1158 ng/g. Au was detected also in kidneys and spleen, which accumulated around 3000 ng Au/g. 41164

DOI: 10.1021/acsami.7b13100 ACS Appl. Mater. Interfaces 2017, 9, 41159−41167

Research Article

ACS Applied Materials & Interfaces

different stains used for each analysis. For apoptosis and necrosis we used a Sigma Annexin V-FITC apoptosis detection kit, whereas for live−dead analysis we used freshly prepared PI solution. Noteworthy, in cells that were not subjected to EM field (Figure 4C), the percentage of apoptotic cells remained significantly lower, around 7% of the total number of cells.

4. CONCLUSIONS Here we demonstrated the cytotoxic potential of