The MWCNT-7 material were a gift from Morinobu Endo and Shuji Tsuruoka (Shinshu University, Nagano, Japan), obtained through the Mitsui & Co., Ltd. (MWCNT-7, lot #05072001 K28) and previously characterized for length and diameter . The MWCNT-HT and ND material are derivate of the MWCNT-7 material and were a gift from Mauricio Terrones (Pennsylvania State University, College Park, PA).
Length and diameter
A protocol was established for the measurements of diameter and length of raw MWCNT material using an SEM/Scanning Transmission Electron Microscope (STEM, S-5500 ultrahigh resolution SEM with STEM capabilities, Hitachi High Technologies America Inc., Schaumburg, IL 60173). The samples were prepared by adding a portion of the raw MWCNT material into a glass vial with isopropanol and sonicating for approximately 5 min to produce a well dispersed sample. A TEM grid (200 mesh Cu grid coated with carbon, SPI Supplies, West Chester, PA) was then dipped into the suspension and allowed to air dry. Without further coating, the material was examined by SEM. Initially a low magnification was used (~ 5-10kX) to locate fibers to measure. Once a fiber was located the magnification was increased appropriately to measure width and length. Using the measuring tools of the electron microscope’s software, straight lines were manually drawn to connect the desired distances to be measured. For length measurements, the longest straight line was drawn between two extremities of a fiber without following the curvatures of the fiber. For the width, measurements were taken by drawing a straight line of the distance perpendicular to the fiber’s walls. A minimum of 200 individual MWCNT structures were measured for each sample.
Scanning transmission electron microscopy (STEM) with energy dispersive X-ray spectroscopy (EDS) was used to qualitatively assess the purity of the three MWCNT materials by identifying the presence or absence of residual catalyst material in the material. Bright-field and dark-field electron microscopy were used to identify the catalyst material and EDS was used to confirm the elemental composition. High-resolution bright-field images were collected at an accelerating voltage of 200 kV (Hitachi HD-2300A STEM, Hitachi High Technologies America, Schaumburg, IL 60173). EDS spectra were collected to confirm the presence of Fe-rich catalyst material (Bruker Quantax, Bruker Nano Analytics, 12,489 Berlin, Germany). Inductively coupled plasma mass spectrometry (ICP-MS) was used to quantitatively measure residual metal contaminants in each MWCNT material. MWCNT samples were prepared and digested in triplicate. Dry MWCNT material was suspended in H20 at a concentration of 1.0 mg/mL, vortexed for 10 s, and aliquoted (200 μL) into 40 mL PFTE digestion tubes containing 9 mL of ultrapure HNO3 and 1 mL of ultrapure H2O2. Tubes were capped and microwave-digested for 30 min at 200o C. After cooling, digested samples were decanted into 50 mL polycarbonate centrifuge tubes, the digestion tubes were rinsed twice with 10 mL of H2O, and the rinses were added to the 5 mL polycarbonate tubes. The volume was adjusted to 40 mL with H2O. Samples were analyzed for 52Cr, 58Fe, 62Ni, and 59Co against certified reference standards using a Nexion inductively-coupled plasma mass spectrometer (Perkin-Elmer, Norwalk CT).
Hydrodynamic diameter (DH) of each material was measured using photon correlation spectroscopy (PCS). Zeta potential was determined for each material suspended in water using laser Doppler electrophoresis . All measurements were performed at 25 °C with a 633 nm laser at a 90° scattering angle (Zeta-sizer Nano ZS90, Malvern Instruments, Worcestershire, UK). The analyses were performed assuming a medium refractive index of 1.332, viscosity of 0.890 cP, dielectric constant of 78.3, and Smoluchowski approximation, f(κa) value of 1.5. Each suspension of MWCNT material was subject to ultrasonic agitation using a probe sonicator (XL 2000, QSonica, Newtown, CT) fitted with a 3-mm titanium probe tip. The delivered energy, as verified calorimetrically , was 27,600 J per sample. Distilled and deionized water that was passed through a 0.025 μm pore-size membrane (Anotop 25, Whatman International Ltd., Maidstone, England) was used to dilute each sample prior to analysis. Suspension stability index (SSI) analysis was conducted on each particle for each cell culture medium based on previously described methods . Briefly, MWCNT suspensions in DM were sonicated, diluted to 0.1 mg/mL in each culture medium, immediately placed into cuvettes, and were assayed for absorbance at 325 nm, 500, and 550 nm on an Evolution 300 spectrophotometer with VisionPro software (ThermoScientific) at each hour over a 48 h period. Medium blanks were used to correct absorbance values for changes in absorbance over time for each medium. Three independent experiments were run. All data were normalized to 1 at zero hr. Quartic curve plots and regression analyses were conducted in SAS JMP v13.2. Parameter comparison and curve equivalence analyses (α = 0.05) were conducted to determine differences in SSI over time between MWCNT particles.
Two pulmonary epithelial cell types were used in the present study. All cells were maintained at 37 °C and 5% CO2 with standard aseptic procedures. Immortalized human bronchial epithelial cells (BEAS-2B, ATCC, Manassas, VA) of less than 10 passages upon arriving in our laboratory were used to examine cytotoxicity, nuclear uptake, cell cycle arrest, mitotic aberrations, centrosome integrity, and spindle pole integrity. BEAS-2B were cultured in Dulbecco’s Modified Eagle Medium (DMEM) media supplemented with 10% (v/v) serum (Invitrogen, Grand Island, NY) and 1% (v/v) antibiotic-antimycotic (Corning, Corning, NY). Primary small airway respiratory epithelial cells (SAEC; Lonza, Walkersville, MD) from a non-smoking human donor were used to examine cytotoxicity, nuclear uptake, cell cycle arrest, aneuploidy, and clonal growth. The normal karyotype of the primary cells was essential for the examination of aneuploidy. The SAEC were cultured following manufacturer’s directions and using Cabrex media (Lonza, Walkersville, MD). Epithelial phenotype was identified in both cell types through EM analysis of stained cytokeratin 8 and 18 (data not shown) .
Preparation of materials
Stock MWCNT material was subjected to 4–6 h of ultrasonic agitation over ice using a 3 mm titanium probe tip sonicator (Sonics and materials, Inc., Newtown, CT) set to 8 kHz for even dispersion in water. Just prior to use, the stock suspensions were dispersed similarly for one minute with a 10 s pulse in order to avoid an increase in temperature. Additionally, the media suspension containing the appropriate volume of stock MWCNT material was sonicated for 10 s before application to cell surface. Vanadium pentoxide (V2O5, sigma, St. Louis, MO) was suspended in dH2O and sonicated in a water bath (Branson 2510, fisher, Pittsburgh, PA) cooled with ice for 30 min immediately prior to addition to culture media. Sodium arsenite (arsenic, sigma, St. Louis, MO) was dissolved in dH2O
The BEAS-2B and SAEC were seeded in parallel culture dishes according to assay protocol. MWCNT doses were based on mass per volume of culture media (μg/mL) and also reported as mass per culture surface area (μg/cm2). Cells were exposed to MWCNT material suspended in appropriate culture media for either 24 or 72 h depending on assay requirements. Three independent experiments were performed for each assay.
MWCNT material in cell nuclei
Confocal Raman spectroscopy was used to determine nuclear uptake and spatial orientation of each MWCNT material. Both BEAS-2B and SAEC were grown on glass chamber slides until 70% confluence and exposed to 24 μg/mL (4.2 μg/cm2) MWCNT-7, HT or ND for 24 h. After exposure cells were washed twice with phosphate buffered saline (PBS), fixed with 100% ethanol, and analyzed. The spectra of the MWCNT reference materials were generated using a Horiba LabRAM HR (Horiba Instruments, Edison, NJ, USA), equipped with an optical microscope, a 1024 × 256 pixel, Synapse CCD detector, a 600 grooves/mm grating, and a 473 nm argon laser. The parameters used to obtain the spectral data were as follows: 100 μm pinhole, 100x objective, a neutral density filter that attenuated all but 10% of the laser power, which resulted in a laser power at the sample of 286 μW, and two accumulations of each spectrum, collected for 5 s each, that were averaged together. A Raman map was generated to permit the analysis of a larger area containing the MWCNT material, and the resultant spectral data acquired from twenty different locations was baseline-corrected and averaged using the Horiba LabSpec 6 software package.
Cells were identified through brightfield imaging and Raman mapping of the cells was performed using a classical least squares (CLS) analysis for silica (glass slide), cellular protein, and MWCNT material using basis spectra. 3D renderings were produced using this data to determine the distribution of the MWCNT material within the nucleus. Raman spectroscopy was performed using an exposure time set to 1 s × 2 accumulations per pixel. The mapped areas were approximately 50x50x10 (XxYxZ) μm with a mapping step size set to 1 μm. Horiba LabSpec v6 software was used for data reduction and analysis.
Enhanced dark-field light microscopy was used to measure the partitioning of MWCNT into the nucleus of BEAS-2B cells. MWCNT have dimensions less than the wavelength of light, have closely packed atoms, and typically have a refractive index significantly different from that of biologic tissues and/or mounting medium. These factors cause nanoparticles, in general, and MWCNT, specifically, to be efficient light scattering structures. The enhanced dark-field microscope images only the light scattered by structures in the optical path. Typical biologic tissues such as cells and cell nuclei, and even the mounting media, produce minimal scattered light, and produce images in the enhanced dark-field microscope which are orders of magnitude lower in intensity than MWCNT. These characteristics which produce significantly greater scattering of light by nanoparticles produce images in which nanoparticles stand out with large, bright intensity compared to the surrounding biologic tissues that do not significantly scatter light. The significant intensity of scattered light from nanoparticles imaged by the enhanced dark-field microscope also produces a bright envelope or halo of scattered light about the outer edges of the nanoparticles which is significantly brighter than adjacent tissue that does not scatter light. Because of the significant scattering of light by nanoparticles, the enhanced dark-field microscope is able to detect nanoparticles in tissues and sections which could not otherwise be detected by a standard light microscope. The theory and application of this microscopy method to detect numerous types of nanoparticles (particles with dimensions less than 100 nm) in a variety of nanoparticle studies are described in detailed elsewhere .
For enhanced dark-field microscopy in the present study, the cells were grown to 70% confluence on glass chamber slides (Nunc™ Lab-Tek™ II, Waltham, MA), serum starved for 24 h, and exposed to 0.024, 0.24, 2.4 and 24 μg/mL (0.0042, 0.042, 0.42 and 4.2 μg/cm2) of each MWCNT material for 24 h. Cells were washed twice with PBS and fixed with 100% ice cold methanol. After fixation, nuclear content was fluorescently stained with DAPI (Vectashield, Vector Laboratories, Burlingame, CA) and individual nanotubes were counted using a high signal-to-noise, enhanced dark-field based illumination optics adapted to an Olympus bX-41 microscope (CytoViva, Auburn, AL 36830). Cells over five slides were counted per treatment. For counting of the partitioning of MWCNT into nuclei each cell nucleus of the slide was examined under enhanced dark-field illumination and any associated individual nanotubes were identified. Each individual nanotubes was then examined to determine if it was potentially partitioned within the cell nucleus by focusing from the top of the cell nucleus to the bottom. To be considered as being within the cell nucleus, the focusing over the entire cell nucleus had to demonstrate that the optical section(s) imaging the nanotube in the nucleus were sandwiched by optical slices above and below that were purely of cell nucleus and did not contain the individual nanotubes. The microscope was then re-focused on the nanotube in the nucleus and the microscope illumination was switched to fluorescent illumination of the DAPI nuclear stain. The nuclear partitioning of the individual nanotube was confirmed by the absence of the DAPI staining of nuclear material where the nanotube was located. MWCNT contained within the cell nucleus was reported as a mean ± SD of individual MWCNT in the nucleus per 1000 nuclei.
The presence of MWCNT material within the nucleus was confirmed by analyzing BEAS-2B cells exposed to 2.4 μg/mL MWCNT-7 for 24 h via transmission electron microscopy (TEM). The methods used for TEM sample preparation were similar to those previously followed . Briefly, cells were fixed in 2% glutaraldehyde in sodium phosphate buffer, pH 7.2, for 2 h, post-fixed in osmium tetroxide, dehydrated through an ethanol series, and embedded in Spurr’s resin (Sigma, St Louis, MO). Silver-gold sections were stained in 2% aqueous uranyl acetate and Reynolds’ lead citrate, observed using a JEOL 1200 EX electron microscope, and recorded digitally.
Cytotoxicity of each MWCNT material was measured in both cell types. Cells were seeded in flat-bottom 96 well plates (Becton Dickinson Franklin Lakes, NJ) and exposed to 0.024, 0.24, 2.4 and 24 μg/mL (0.015, 0.15, 1.5 or 15 μg/cm2) of each MWCNT material for either 24 or 72 h. A 0.316 or 3.16 μg/mL or (0.2 or 2 μg/cm2) dose of V2O5 was used as positive control in the BEAS-2B or SAEC, respectively. Each treatment was measured in triplicate and the assay was repeated three times for each cell type. Cytotoxicity was assessed using the alamarBlue cell viability assay protocol following manufacturer directions (Invitrogen, Carlsbad, CA). Fluorescence was measured using a fluorescent spectrophotometer (LS50B, Perkin Elmer, Bridgeville, PA) with a 570 nm excitation and 585 nm emission wavelength. The fluorescence intensity was measured for each well. Cell viability is equivalent to a reduction in fluorescence intensity and was reported as a mean ± SD across all three experiments normalized to control.
Laser scanning fluorescent confocal microscopy with differential interference contrast was used to analyze mitotic aberrations after exposure to each MWCNT material in the BEAS-2B cell (LSM 710, Carl Zeiss MicroImaging Inc.,Thornwood, NY). The relatively high mitotic index of the BEAS-2B cell type allows for sufficient examination of dividing cells. Cells were seeded on glass chamber slides (Nunc™ Lab-Tek™ II, Waltham, MA) until 70% confluence and exposed to 0.024, 0.24, 2.4 and 24 μg/mL (0.0042, 0.042, 0.42 and 4.2 μg/cm2) of each MWCNT material or 0.316 μg/mL (0.06 μg/cm2) V2O5 for 24 h.
After exposure, cells were washed twice and fixed with 100% methanol at 4 °C (Fisher Scientific, Waltham, MA). Dual chambers were prepared for each dose. The cells were stained for mitotic aberration analysis via fluorescent labeling of the DNA and immunofluorescent labeling of the mitotic spindle and centrosomes. The DNA was fluorescently labeled using DAPI (Vectashield, Vector Laboratories, Burlingame, CA). The β-tubulin of the mitotic spindle was labeled using a rabbit anti-β-tubulin primary antibody (Abcam, La Jolla, CA, USA) and goat anti-rabbit IgG secondary antibody conjugated with rhodamine red (Invitrogen, Carlsbad, CA). The centrosomes were labeled using mouse anti-pericentrin primary antibody (Covance, Austin, TX, USA) and goat anti-mouse IgG antibody conjugated with Alexa 488 (Invitrogen, Carlsbad, CA). Cells were examined and divisions were analyzed by photographing serial slices through the z-plane based on the depth of the cell and optical properties of the stain (Zen, Carl Zeiss MicroImaging Inc., Thornwood, NY). A minimum of 50 mitotic cells of good centrosome and mitotic spindle morphology were analyzed for each dose. Three independent experiments were conducted for a total of 150 cells.
Quantitative analysis of aberrant mitoses was based on spindle morphology; a mitotic cell with monopolar or multipolar spindle morphology was considered aberrant. Aberration was reported as a percentage of total mitotic cells analyzed for each dose across all three experiments. Centrosome and spindle pole integrity were assessed quantitatively. The association between MWCNT material and the labeled nuclear structures was also examined qualitatively by overlaying the fluorescent images with the differential interference contrast filter. The mitotic index is equivalent to the percentage of mitotic divisions in 100 cells per treatment.
Laser scanning confocal fluorescent microscopy imaging of fluorescently-labeled centromere of chromosomes 1 & 4 in SAEC was used to determine aneuploidy and centromere fragmentation after exposure to each MWCNT (LSM 710, Carl Zeiss MicroImaging Inc., Thornwood, NY). Chromosomes 1 & 4 were pragmatically chosen due to their size and labeling efficiency. Cells were seeded on glass chamber slides (Nunc™ Lab-Tek™ II, Waltham, MA) until 70% confluence and exposed to 0.024, 0.24, 2.4 and 24 μg/mL (0.0042, 0.042, 0.42 and 4.2 μg/cm2) of each MWCNT material or 3.16 μg/mL (0.06 μg/cm2) V2O5 for 24 h. After exposure, cells were washed twice and fixed with a 3:1 (v/v) mixture of methanol and acetic acid (Fisher Scientific, Waltham, MA). Chromosomes 1 and 4 were labeled via fluorescence in situ hybridization (FISH) of centromeric DNA (Abbott Molecular, Des Plaines, IL) and cells were fluorescently counterstained with DAPI (Vectashield, Vector Laboratories, Burlingame, CA) for chromatin content. Each batch of FISH probes were tested on normal human metaphase spreads isolated from normal human lymphocytes for a bright FISH signal at the correct location on the correct chromosome. The SAEC cell type has a normal karyotype, therefore chromosome enumeration for quantitative analysis of aneuploidy is possible. Cells were examined and scored according to the most stringent guidelines available ACMG . Cells with three or greater than four signals for either chromosome were recorded as a gain; cells with less than two signals of either chromosome were recorded as a loss. Cells with exactly four signals were considered in active synthesis and excluded from the analysis. Only cells with good morphology were included in the analysis; cells with obvious signs of necrosis, apoptosis, and decondensed centromeres were not scored . Use of centromeric probes allowed for a quantitative analysis of aneuploidy, centromere fragmentation, and translocations between chromosomes 1 and 4. Insertions between these two chromosomes, an extremely rare event, were noted but not quantitated. A fragment was determined if it was 1/3 or less the size of the normal centromere signal within that same cell. Translocations in interphase cells are identified by the presence of two fluorescent signals less than one signal distance apart. Only signals less than one signal distance apart that also had overlapping pixels were labeled positive for insertion/translocation, accordingly. A minimum of 100 interphase cells with satisfactory fluorescent signal were analyzed for each dose. Three independent experiments were performed for a total of at least 300 cells included in the analysis. Slides were coded and scored by three independent investigators. Aneuploidy was reported as a mean ± SD of the percentage of cells with either a gain or loss for each dose across all three experiments.
Cell cycle analysis
Bivariate flow cytometry using the Click-iT EdU Alexa Fluor 647 flow cytometry assay kit (Molecular Probes, Eugene, OR) with 7-aminoactinomycin D (7-AAD, Invitrogen, Carlsbad, CA) allows for a more accurate analysis of the cell cycle compared to single-color methods. EdU (5-ethynyl-2′-deoxyuridine), a nucleoside analog of thymidine, is incorporated into DNA during the S phase of the cell cycle and covalently-labeled with Alexa Fluor 647 via a click chemistry reaction between an azide in the fluorophore and an alkyne within the EdU. The 7-AAD fluorophore is incorporated into the DNA of all fixed cells thereby staining for G1 and G2 phases of the cell cycle.
BEAS-2B and SAEC cell types were seeded in T25 flasks (Falcon, Corning, NY) until 70% confluence. BEAS-2B cells were exposed to 24 μg/mL (2.88 μg/cm2) of each MWCNT material and 5 μM arsenic for 24 h. EdU was applied after 22 h of exposure to allow for incorporation into the DNA. Cells were washed twice with PBS (Gibco, Waltham, MA) and 0.25% (v/v) trypsin in EDTA (Gibco, Waltham, MA) was used to remove cells from the flask surface. Two exposures were analyzed for the SAEC cell type requiring separate methods. First, SAEC cells were exposed to 10 μM arsenic or 24 μg/mL (2.88 μg/cm2) MWCNT material for 24 h. EdU was applied after 12 h of exposure. Cells were washed twice with PBS and removed from the flask with 0.25% (v/v) trypsin in EDTA (Lonza, Basel, Switzerland). Second, SAEC cells were exposed to 10 μM arsenic or 2.4 μg/mL (0.288 μg/cm2) MWCNT material for 72 h. EdU was applied after 12 h of exposure. Cells were washed twice with PBS and fresh media was applied for a 24 h recovery period. Each treatment was performed in triplicate. Cells were stained according to manufacturer’s instructions and run through a flow cytometer (LSR II, BD Biosciences Immunocytometry Systems, San Jose, CA). Ten thousand events were collected and the dual-labeled fluorescent DNA content was analyzed (FlowJo v10, FlowJo, Ashland, OR). Gating was set to exclude debris, non-cellular material, and doublets. The percentage of cells in G1, S, and G2 phases of the cell cycle were determined via manual gating of the bivariate analysis of the two fluorescent signals and reported as a mean ± SD across all experiments.
Enumeration of SAEC colonies was used to determine the clonal growth after exposure to each MWCNT. Cells were seeded in T25 flasks (Falcon, Corning, NY) until 70% confluence and exposed to 0.024, 0.24, 2.4 and 24 μg/mL (0.00288, 0.0288, 0.288, and 2.88 μg/cm2) of each MWCNT material or 3.16 μg/mL (0.4 μg/cm2) V2O5 for 24 h. After exposure cells were washed twice and removed from the flask surface with 0.25% (v/v) trypsin in EDTA (Lonza, Basel, Switzerland). Cells were reseeded in 6-well flat bottom plates (Falcon, Corning, NY) at 500 cells/well to allow for clonal growth from a single cell. Colonies were grown for one month and stained with a 10% (v/v) solution of crystal violet in neutral buffered formalin (Sigma, Saint Louis, MO) to preserve and identify clonal morphology. A stereo microscope (SZX12, Olympus, Shinjuku, Japan) was used to count the colonies in each of the six wells. The mean ± SD of colonies was calculated and reported as a percentage of control.
All analyses were performed using SAS/STAT (Version 9.4) for Windows, and JMP version 12 (SAS Institute, Cary NC). Data were analyzed using appropriate linear models including one-way and two-way analysis of variance (ANOVA). Some experiments were performed using a randomized blocks design and mixed-model ANOVAs were utilized to include block as a random factor. The assumptions of the models such as homogeneity of variance were assessed by inspection of residual plots. For some variables a log transformation was utilized to reduce heterogeneous variances. All differences were considered statistically significant at p < 0.05.