Nanocomposite Synthesis and Characterization:
Fe3O4 core and TiO2 shell NPs (Fe3O4@TiO2) were synthesized through a modified low temperature alkaline hydrolysis method as previously described.(Arora et al. 2012 22158944) Fe3O4 nanoparticle cores were synthesized by stirring a solution of FeCl2 and FeCl3 in 24 mM citric acid for 3 hours at room temperature. The mixture was then allowed to gel in static air at 70°C for 24 hours, forming the Fe3O4 core nanoparticles 1.5 to 3 nm in size. This solution was chilled and stirred vigorously with the gradual addition of chilled TiCl4 in HCl at 4°C, allowing for the Ti shell to form. Elemental concentration determination of nanocomposite suspension was performed by measuring titanium and iron concentrations by ICP-MS, at the Northwestern University Quantitative Bioelemental Imaging Center on an X series II ICP-MS (Thermo scientific, West Palm Beach, FL). The calculation used to determine the molarity of nanocomposites was previously described,(Arora et al. 2012 22158944) using Atomic Force Microscopy (AFM) sizing calculations and elemental concentrations determined by ICP-MS.
Following synthesis, Fe3O4@TiO2 nanocomposites were dialyzed (dialysis tubing pore size = 2,000 MWCO) in 10 mM Na2HPO4 (pH 4.5) and stored at 4°C. Under these conditions phosphate molecules attach to the nanoparticle surface; this preparation constituted “bare” nanocomposites. (Michelmore et al. 2000 WOS:000088087000021)
MIBG (Sigma-Aldrich, St. Louis, MO) was bound with 3,4-dihydroxyphenylacetic acid (DOPAC) through a peptide bond forming reaction using EDC (Thermo Scientific), following the manufacturer’s instructions. The final concentration of MIBG-DOPAC was 2.34 mM. DOPAC was used as a linker because it has a high affinity for the surface of nano-sized TiO2 being a catechol. DOPAC also has a carboxyl group that can form a peptide bond with the amino group of MIBG.(Creutz and Chou 2008 18366179; Paunesku et al. 2007 17274661; Thurn et al. 2009 19242946) Fe3O4@TiO2 nanocomposites were mixed overnight with DOPAC-MIBG in an oxygen free atmosphere; the resultant nanoconjugates were dialyzed in 10 mM sodium phosphate buffer (dialysis pore size of 2,000 MWCO). Using a calculation approach previously described(Arora et al. 2012 22158944) we estimated that mixing DOPAC-MIBG solution as prepared V:V with 22.3 μM Fe3O4@TiO2 nanocomposites lead to DOPAC-MIBG covering roughly 70% of the nanoparticle surface.
DOPAC-Fe3O4@TiO2 nanoconjugates were prepared by combining equal volumes of 7.838 mM DOPAC with 28.93 μM Fe3O4@TiO2 nanocomposites in an oxygen-free environment followed by mixing the conjugation reaction overnight at 4°C. DOPAC-Fe3O4@TiO2 nanocomposites were then dialyzed for 2 hours in 10 mM Na2HPO4 buffer. We estimated that under these circumstances molecules of DOPAC covered roughly 100% of the nanoparticle surface.
Elemental makeup and the shape of Fe3O4@TiO2 -MIBG nanoconjugates were evaluated by EDS-STEM which was performed at Northwestern University’s Atomic and Nanoscale Characterization Experimental Center (Supplemental Figure 1a-c). In preparation for EDS-STEM, Fe3O4@TiO2 -MIBG nanoconjugates were diluted 1:100 in ddH2O, drop cast onto 150 square mesh copper grids with a carbon film support, allowed to dry, and then imaged on a Hitachi HD-2300 Dual EDS Cryo STEM (Supplemental Figure 1a).
Infrared determination of nanocomposite coating (Supplemental Figure 2) was performed using infrared spectroscopy at the Infrared Environmental Imaging (IRENI) instrument at University of Wisconsin Synchrotron Center. Droplets of different nanoconjugate colloids or component solutions were cast and dried on Ultralene membrane supports and scanned in 2D. Chemograms for the areas of interest were obtained. Addition of each new nanoconjugate coating can be followed by appearance of new spectral features (Supplemental Figure 2).
Zeta potential measurements of nanoconjugates were also obtained (Supplemental Table 1) using a protocol that was adapted from the procedure recommended by the Nanotechnology Characterization Laboratory at the National Cancer Institute.(Clogston 2009) Bare nanocomposites, DOPAC-Fe3O4@TiO2, and MIBG-Fe3O4@TiO2 nanoconjugates were diluted 1:100 in 10 mM filtered NaCl using the following constants (temp: 25°C, viscosity: 0.891, dielectric constant: 78.6, Henry function: 1.5, refractive index: 1.33) on a Zeta sizer Nano (Malvern, Worcestershire, United Kingdom). Fe3O4@TiO2-MIBG nanoconjugates have a mean ZP of -40.887± 1.85, bare (phosphate covered) Fe3O4@TiO2 nanoconjugates have a mean ZP of -37.1± 1.91, and DOPAC coated nanocomposites have a mean ZP of -33.367 ± 0.71.
Nanosight measurements were also performed using a Nanosight LM10-HS (Malvern, Worcestershire, United Kingdom) in lieu of dynamic light scattering (Supplemental Table 2). Because of the polydispersity of nanocomposites, presence of nanocomposite aggregates and the non-spherical shape of these objects, these results are not as reliable a source of nanoparticle size information as the AFM or TEM data.
To further confirm the spatial and aggregation characteristics of different nanocomposites, Cryo-TEM (Figure 1a-c) was performed at the Northwestern University Biological Imaging Facility. Nanocomposite 1:100 dilutions in full media (DMEM + 10% FBS) of bare (phosphate coated) Fe3O4@TiO2 nanocomposites, DOPAC-Fe3O4@TiO2 nanoconjugates, and MIBG-Fe3O4@TiO2 nanoconjugates were drop cast on plasma treated lacy carbon TEM grids, plunge frozen in liquid ethane using a FEI Vitrobot Mark IV, and Cryo-TEM was performed using a JEOL 1230 TEM at 120 kV. Image brightness levels were adjusted to enhance contrast. The spatial and aggregation properties of different nanocomposites are similar, close to value obtained by AFM and smaller than the size estimates obtained from the light scattering measurements done with the Nanosight instrument.
EGFR B-loop peptide (DOPAC-MYIEALDKYAC-COOH) and scrambled peptide (DOPAC-EAKLDYMCIYA-COOH) were synthesized by the IBNAM (now The Simpson Querrey Institute for Bionanotechnology) Core Facility of Northwestern University’s Institute for Bionanotechnology in Medicine. The DOPAC group at the N-terminus of the peptide served as a linker to conjugate the peptide to the TiO2 surface of the nanoparticles. B-loop peptide was dissolved to a concentration of 700 μM in ddH2O bubbled with N2, then mixed V:V with 22.3 μM MIBG-nanocomposites or 28.93 μM bare nanocomposites. A separate set of nanocomposites was conjugated to a scrambled peptide. The conjugation was performed in an oxygen free atmosphere overnight at 4°C. At a concentration of 350 μM peptide and 11.15 μM MIBG-Fe3O4@TiO2, it was estimated that the B-loop peptide should cover roughly 23% of the MIBG-nanoparticle surface, while with bare nanocomposites the surface coverage was estimated to be roughly 18%(Arora et al. 2012 22158944). Experiments with cells in culture were performed within 4 hours following conjugation.
Cell culture
Neuroblastoma cell lines SK-N-AS and SK-N-DZ and cervical cancer cell line HeLa were purchased from American Type Culture Collection (ATCC, Manassas, Virginia). These cells were grown in DMEM supplemented with 10% non-heat inactivated FBS with penicillin/streptomycin and non-essential amino acids at 37°C and 5% CO2. Neuroblastoma cell lines NBL-W/S and NBL-W/N were a generous gift from S.L. Cohn (Department of Pediatrics, University of Chicago, Chicago, IL). These cells were grown in RPMI-1640 medium supplemented with 10% heat inactivated FBS with penicillin/streptomycin and L-glutamine.
Cell viability assay
SK-N-AS, SK-N-DZ, NBL-W/S, NBL-W/N, and HeLa cells were trypsinized, collected, counted, and plated (5-6 x 103 SK-N-AS cells per well, 6-7 x 103 SK-N-DZ, 9 x 102 NBL-W/S, 8-9 x 103 NBL-W/N, or 2.5 x 103 HeLa) into 96-well plates and allowed to attach overnight. Five to six wells for each cell line were used as biological replicates in each experiment. Cells were treated with varying concentrations of bare Fe3O4@TiO2 nanocomposites, DOPAC-Fe3O4@TiO2 nanoconjugates, MIBG-Fe3O4@TiO2 nanoconjugates, free DOPAC, free MIBG, DMSO (control for free MIBG experiments), or Na2HPO4 buffer (10 mM, vehicle control for nanocomposites). In radiosensitization experiments, cells were treated with nanocomposites or nanonconjugates for 1 hour, followed by varying doses of ionizing radiation. Cesium source (662 keV) Gamma Irradiator (Gammacell 40, Atomic Energy of Canada Ltd.) was used under the supervision of the Office of Research Safety, Health Physics Services, Northwestern University.
Nanocomposite/nanoconjugate and irradiation treatments were performed in complete media over a period of 72 hours. Equivalent “blank” wells without cells, but with an identical concentration of treatment reagents were used as blank controls (n=5), in order to account for any possible modification of absorbance readout that could occur because of the treatment materials used. After incubation, a tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt] (MTS reagent, Promega, Madison WI) was added as 1/10th of the volume into each well, and the 96 well plates were incubated an additional 2-4 hours at 37°C. Initially, treatment media was removed from both treatment and blank wells before adding MTS and new media; this approach was discontinued when no significant differences in results were observed when MTS was added to wells without any prior manipulation. The latter approach was used subsequently in order to mitigate the risk of inadvertent removal of less adherent cells. Absorbance readings were measured using a SpectraMax M5 (Molecular Devices, Sunnyvale, CA) plate reader at 490 nm. Absorbance values measured for “blank” controls were averaged and subtracted from the treatment values. Resultant absorbance value was divided by the average of the absorbance values for each respective negative control, providing percentage of cell viability as a function of the control (surviving fraction). For radiation sensitization determination, each adjusted absorbance value was divided by the average of each treatment group respective baseline (cells treated with nanocomposites or nanoconjugates but not exposed to radiation) absorbance, to obtain an adjusted percentage of survival, to account for baseline cell death.
Evaluation of nanocomposite and nanoconjugate uptake by ICP-MS
SK-N-AS and SK-N-DZ cells were trypsinized, counted, and plated (5-8 x 105 of SK-N-AS cells, or 6-10 x 105 SK-N-DZ cells) onto 6-well plates and allowed to attach overnight. Empty wells with identical treatment conditions were also prepared, to act as a control for potential artifacts such as adherence of nanocomposites to the bottom of the wells in absence of cells. Additional wells seeded with an identical number of SK-N-AS or SK-N-DZ cells were used to estimate the final number of cells per well at the conclusion of experiment and to determine background elemental concentrations. Treatments with 250 nM bare Fe3O4@TiO2 nanocomposites, 250 nM DOPAC-Fe3O4@TiO2 nanocomposites, or 250nM MIBG-Fe3O4@TiO2 nanoconjugates were done in compete media (total volume of media per well was 1 ml) for 1 hour at 37°C; three wells per treatment represented biological replicates. After treatment, the wells were washed 1-3 times with PBS and once with acidic glycine. Finally, 500 μL of 70% HNO3 (re-distilled, >99.999% trace metal basis) was added per well and cells and nanoconjugates were digested for 2 hours at room temperature. Samples were then transferred into 15ml metal-free Falcon tubes, mixed with 10 mL 3% HNO3 in ddH2O containing 3 ppb 115In (as an internal control), and allowed to digest additionally overnight at 70°C. Samples were evaluated for elemental concentrations of 47Ti, 57Fe, and 115In using an X series II ICP-MS. Ti concentration was used as a proxy for nanoparticle/nanoconjugate concentration. Average background elemental quantity obtained from cell free nanocomposite treated blank wells was subtracted from each test sample, to arrive at a final total concentration of nanocomposites taken up by cells. This number was divided by the “end of experiment” cell count to arrive at a Ti concentration per 105 cells. Uptake of bare nanocomposites was used as a standard and uptake of nanoconjugates was expressed as a percentage of the Ti concentration found for uncoated nanocomposites.
Evaluation of Apoptosis/Necrosis by Flow Cytometry
SK-N-AS cells were seeded onto 6-well plates (2 x105 cells per well) and allowed to attach overnight. Cells treated with 250 nM DOPAC-Fe3O4@TiO2 nanocomposites for one hour as well as untreated controls were either sham irradiated or exposed to 10 Gy of gamma rays from a 137Cs irradiator. Cells were then incubated for 24 hours, washed with PBS, trypsinized and processed according to manufacturer’s instructions. 105 cells in 100 μL of annexin binding buffer were incubated with Annexin-FITC and Propidium Iodide (PI) for 15 minutes at room temperature, placed on ice, and immediately evaluated by flow cytometry using a BD LSR Fortessa Analyzer (BD Biosciences, Franklin Lakes, NJ) at the Robert H. Lurie Comprehensive Cancer Center Flow Cytometry Core Facility at Northwestern University. For each biological replicate, 5000 gated events were analyzed.
Evaluation of 53BP1 Foci
A total of 200,000 SK-N-AS cells were seeded onto barrier slides and allowed to attach overnight. Cells in 1 ml full medium were then treated with varying concentrations of DOPAC-Fe3O4@TiO2 nanoconjugates or bare nanocomposites for 1 hour. Cells were then irradiated with 2 Gy and incubated for 4 hours. Slides were washed with PBS and cells fixed in 3.6% formaldehyde for 10 minutes at room temperature. Cells were then permeabilized with PBS-Triton (0.2%) for 10 min, rinsed three times with PBS-BSA (1%) – Tween (0.5%) and processed further in the same buffer. Slides were incubated for 1h with primary antibody against 53BP1 (ab21083 – Abcam, Cambridge, UK) used at 1:200 dilution, washed and incubated for 45 minutes with fluorescent secondary antibody (Alexa Fluor 488 – Goat Anti-Rabbit, ab150077, Abcam, Cambridge UK). Nuclei were counterstained with propidium iodide (2.5 μg/ml). Cells were imaged at 40x magnification with a full field fluorescent Zeiss microscope equipped with a CoolSNAP EZ CCD camera (Photometrics, Tucson AZ, US).
Four experiments were performed with two replicate slides for each treatment condition conducted. The 53BP3 foci in each replicate were counted by a different researcher. Multiple images of each slide were taken, and foci present in at least 100 cells were counted for each treatment group. All experiments were pooled as indicated (slide numbers ranged between 3 and 7) and statistics generated from the pooled data.
Cryogenic X-ray fluorescence microscopy (Cryo-XFM)
SK-N-AS and SK-N-DZ cells were seeded on 1.5 mm x 1.5 mm Si3N4windows overnight (Silson, UK), then treated either with 4.24 µM MIBG (with a resultant 0.05% DMSO concentration), 25.44 µM MIBG (final 0.30% DMSO), 60 µM MIBG (final 2% DMSO), or DMSO control (0.30% DMSO) in 50 µL of full media for 90 minutes. In nanocomposite treatment experiments, cells were treated with MIBG-Fe3O4@TiO2-B-Loop nanoconjugates (carrying an equivalent of 60 µM of MIBG) for 90 minutes in serum-free DMEM. In addition, another set of cells grown on Si3N4 windows were exposed to 250 nM MIBG-Fe3O4@TiO2, 250 nM DOPAC-Fe3O4@TiO2, or 250nM of bare (phosphate covered) Fe3O4@TiO2 nanocomposites for 60 minutes in 50 µL of full media. The windows were washed twice in a Tris glucose buffer (261 mM glucose, 9 mM acetic acid, 10 mM Tris buffer, pH 7.4) and plunge frozen in liquid ethane using a FEI Vitrobot Mark IV. Frozen hydrated cells were imaged with visible light on a Nikon microscope equipped with an Instec CLM77KCryo-LM stage in order to evaluate the quality of each sample with regard to ice accumulation as well as cell density and distribution.
X-ray fluorescence imaging was done with several different instruments under different conditions. A beam spot size of about 300 or 600 nm was used at the sector 2-ID-D at APS at ANL in combination with a cryo-jet; while the Bionanoprobe at sector 21 LS-CAT was used with a beam spot size 85 nm, and the samples were maintained in vacuum at liquid nitrogen temperature. High-resolution elemental maps were obtained at different angles, allowing subsequent tomographic reconstruction. A monochromatic 10 keV X-ray beam was used and the cells were scanned in “continuous” (fly-scan) mode. Step scans for an area of interest were also done at a step size of 80 nm and per pixel dwell time of 3 sec. To minimize background noise, a Gaussian smoothing filter (σ = 2/3) was applied to the images in the figures presented.
For tomographic reconstruction, scans were done at multiple angles (3-degree increments, total angular range of 138 degrees); reconstructions were performed in Mathematica 9.0 (Wolfram Research, Champaign, IL). In order to circumvent misalignment of particles as a result of sample movement during scanning, the “displacement” of particles with clear projections (Cl, I, K, P and S) was analyzed, and the average shift along the x and y axis of these particles was calculated. Correction of any displacement of the tested particles was achieved by applying this calculated shift to all the particles, hence any misalignment was adjusted. Tomographic reconstruction was attained by implementing a modified Simultaneous Iterative Reconstruction Technique (SIRT) rather than Filtered Back Projection (FBP) since the later yields poor images owing to the limited number of projections. Use of SIRT was described previously with the final visualization of elemental signals of interest using Avizo (FEI, Burlington, MA).(Vo et al. 2014 25320994) Elemental concentration data were extracted for each pixel of the 2D images (elemental quantification, per-pixel fitting) using the MAPS program.(Vogt et al. 2003)
Statistical analyses
All comparisons were performed using student’s T-test at a significance level of 0.05. Data points in all figures correspond to mean +/- standard deviation.