|Home | About | Journals | Submit | Contact Us | Français|
Recently, heat generated by iron oxide nanoparticles (IONP) stimulated by an alternating magnetic field (AMF) has shown promise in the treatment of cancer. To determine the mechanism of nanoparticle-induced cytotoxicity, the physical association of the cancer cells and the nanoparticles must be determined. We have used transmission electron microscopy (TEM) to define the time dependent cellular uptake of intratumorally administered dextran-coated, core-shell configuration IONP having a mean hydrodynamic diameter of 100-130 nm in a murine breast adenocarcinoma cell line (MTG-B) in vivo. Tumors averaging volumes of 115 mm3 were injected with iron oxide nanoparticles. The tumors were then excised and fixed for TEM at time 0.1 to 120 hours post IONP injection. Intracellular uptake of IONP was 5.0, 48.8 and 91.1% uptake at one, two and four hours post-injection of IONP, respectively. This information is essential for the effective use of IONP hyperthermia in cancer treatment.
It is well known that exposing tumor cells to modestly elevated temperatures sensitizes them to chemotherapy and radiation and, depending on temperature and exposure time, decreases their viability  . The first use of magnetic hyperthermia on tumors was reported in 1957, when Gilchrist et al. demonstrated through in vitro experiments that 5 mg of 20-100 nm diameter Fe2O3 nanoparticles in lymph nodes (47 mg of Fe2O3 per gram of tissue) could produce a temperature rise of 14°C in an AMF of 200-240 Oe at 1.2 MHz . Recent advances in nanoparticle technology have allowed for a promising cell-targeted form of therapeutic hyperthermia    .
When appropriately coated, IONP demonstrate excellent biocompatibility with limited toxicity . At the same time, these nanoparticles can be constructed to have a very high specific absorption rate (SAR), resulting in excellent heating properties and targeting capabilities . Considerable work has been done to elucidate the biologic, therapeutic and materials properties of matrix-configuration iron oxide nanoparticles   . This configuration of nanoparticle is typically composed of several, small (3-15 nm) iron oxide crystals distributed within a polymer matrix .
Another configuration of iron oxide nanoparticle, most commonly used in AMF-induced hyperthermia, is the core-shell configuration. In this configuration, larger magnetic iron oxide cores (up to approximately 50 nm) are coated with a polymer shell . This configuration typically demonstrates greater SAR than the matrix-configuration IONP    and, thus, are well-suited to therapeutic hyperthermia applications. To date, few, if any, studies have demonstrated the biodistribution of core-shell IONP . Thus, additional work is necessary to demonstrate the interaction of core-shell IONP with cells and tissues. In this study, we report on the cellular uptake parameters of dextran-coated, iron oxide core-shell nanoparticles.
Theoretical models have been published which have shown that currently available nanoparticles of the concentrations used for nanoparticle hyperthermia will be unable to generate sufficient heating to achieve global hyperthermia . These models assume homogeneous distributions of nanoparticles within a sample; our study suggests that this model is not directly applicably to many biological systems. In addition, inter-particle distance has been shown to be a critical parameter for the generation of heat ; this study, and others, have shown that IONP localize in vacuoles within cells. In addition, studies have demonstrated that iron oxide nanoparticles can be used to achieve global hyperthermia within tissues  . Our data suggest that the distribution of nanoparticles within tumors is not homogeneous. Rather, the nanoparticles aggregate and this aggregation may aid in local heat deposition.
The nanoparticles utilized in this experiment were dextran-coated, iron oxide (magnetite, Fe3O4)-core BNF® nanoparticles with an average hydrodynamic diameter range of 100-130 nm (MicroMod GmBH, Rostock, Germany). The mean magnetite core diameter was approximately 45 nm. The nanoparticle iron concentration was 14.5 mg Fe/ml (33 mg nanoparticle/ml) in deionized water. The synthesis of these nanoparticles was described by Gruettner, et al  and a complete description of their physical characteristics has been published .
A murine breast adenocarcinoma cell line (MTG-B)  was cultured in the Alpha modification of Eagle’s Minimal Essential Medium (MEM, HyClone Laboratories, Inc.) with 1% penicillin/streptomycin (Pen-Strep, HyClone Laboratories, Inc., Logan, UT, USA), 1% L-glutamine (Mediatech, Inc., Manassas, VA), and 10% fetal bovine serum (FBS, Hycolone Laboratories, Inc.) at 37 degrees Celsius in 5% CO2 atmosphere in an incubator (Queue Systems Inc., Parkersburg, VA, USA).
This experiment was approved by Dartmouth’s Institutional Animal Care and Use Committee and all animals were treated humanely, in accordance with the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). To prepare cells for implantation in mice, the cells were exposed to trypsin (0.25% trypsin in EDTA, HyClone Laboratory, Inc.), stained with trypan blue (Hyclone Laboratories, Inc.), counted using a hemocytometer (Fisher Scientific Inc. Pittsburg, PA, USA), and then re-suspended in (serum-free, L-glutamine-free, Pen-Strep-free) Alpha MEM media at a concentration of 107 cells/mL. For each tumor, 100 μL of this solution was injected intradermally into the shaved flanks (one or two tumors per mouse) of female C3H/HE mice (The Jackson Laboratory, Bar Harbor, ME, USA), as described in table 1. Three orthogonal tumor measurements were taken with a digital caliper every two days and the tumor volume was determined using the equation for the volume of an ellipsoid. When tumors reached volumes of greater than 50 mm3 they were considered of appropriate size for analysis.
Nanoparticles (5 mg Fe/cm3 tumor) were injected using a 30-gauge needle (0.34 μL of nanoparticle solution per mm3 of tumor). The tip of the needle was advanced into the center of the tumor and the nanoparticle suspension injected over the course of 30 seconds. The tip of the needle remained in place for five minutes post-injection to optimize nanoparticle distribution. Animals were euthanized and the tumors were excised, sectioned, fixed overnight, and processed for TEM at the pre-determined post-inject time endpoint (table 2).
Tumor tissue samples were fixed in 200 μL of 4% glutaraldehyde (Ted Pella, Inc., Redding, CA, USA) solution overnight and transferred to 200 μL of 0.1 M sodium cacodylate buffer solution (pH 7.4, Ted Pella, Inc.) after three wash cycles with buffer. The samples were prepared for TEM (FEI Company Tecnai F20 FEG TEM operating at 100 kV) at the Dartmouth College Electron Microscopy Facility. Either L.R. White (Polysciences, Inc., Warrington, PA, USA) or Poly/Bed-812 (Polysciences, Inc.) was used as an embedding resin. Samples were stained with 4% osmium tetroxide (Ted Pella, Inc.) and en-bloc stained with 2% uranyl acetate (Ted Pella, Inc.) for one hour, each. Thin sections of 100-110 nm from each tumor sample were cut using a Leica Ultra-Cut Microtome (Leica Microsystems GmbH, Wetzlar, Germany).
To quantify nanoparticle uptake, computer code was written using Matlab 22.214.171.1244 (The Mathworks, Inc., Natick, MA, USA) Image Processing Toolbox. Ten low-magnification (5000x), randomly-chosen TEM fields from each tumor sample were digitally photographed. This magnification allowed for the assessment of multiple cells and associated extracellular spaces within one field of view. Due to electron density similarities—but morphologic differences— of nuclear chromatin and IONP, nuclei were manually excluded from images using a digital computer input tablet (Wacom Technology Corporation, Vancouver, WA, USA). Nanoparticles were never present within nuclei, so this exclusion did not affect quantification accuracy. Artifacts from sample preparation were also excluded. The images of the cells were then manually segmented along their plasma membranes in order to separately quantify internal and external IONP.
Due to the extreme election density of the IONP, we were able to segment each image to identify only IONP in a binary map of the intracellular region of interest (ROI). In the binary ROI, the nanoparticles were given a value of 0 and all other material a value of 1. This ROI map was inverted and summed to count the number of pixels corresponding to intracellular IONP. In a similar fashion, the extracellular space was analyzed using the same grayscale threshold value as the intracellular IONP. Once the pixels corresponding to nanoparticles were determined, an overlay was created to highlight the pixels determined to be nanoparticles and manual confirmation was completed. Using this analysis technique, the ratio of internal vs. external nanoparticles was quantified.
For each tumor sample (three tumors per time point) at times one, two, three, four, five and six hours, the total number of pixels corresponding to intracellular and extracellular nanoparticles was independently quantified. The percentage of intracellular vs. extracelluar IONP was then determined for each sample.
Within the first hour post-injection, 95% of the observed IONP are either associated with the external plasma membrane or within the extracellular space (figure 1). The first indication of significant particle uptake by cells occurs two hours after injection when 48.8 ± 14 % of the nanoparticles are observed within the cells.
Four hours after IONP injection into the tumor, virtually all (91.1 ± 9.6 %) observed nanoparticles are present within tumor cells. The nanoparticle aggregates have become more ordered due to nanoparticle segregation within intra-cytoplasmic vesicles. Figure 2 demonstrates the intermediate time point of three hours. After six hours of incubation, virtually no IONP are observed outside the cells (figure 3). As seen at one day after injection of nanoparticles into the tumors, the nanoparticles are seemingly exclusively contained within intracellular intra-cytoplasmic vesicles (figure 3). The quantification of IONP position for time points one through six hours is presented in figure 4.
These data suggest that virtually all IONP uptake occurs between one and four hours after intratumoral injection. These data suggest that there are two time domains for intratumorally delivered IONP hyperthermia. Dennis et al. demonstrated that aggregated IONP result in improved heating characteristics when compared with non-aggregated IONP . Thus, once tumor cells have aggregated IONP intracellularly, more heat will be deposited into the tumor upon AMF activation with the same field strength and frequency, resulting in greater tumor cytotoxicity.
Multiple TEM sections of each tumor sample are examined. The dimension of the individual TEM sections is 5 mm × 5 mm, at a thickness of 100 nm. While the total volume of tumors sampled with each TEM is relatively small and is assessed in a two dimensional manner, the use of multiple sections over a small region allows for a reasonably accurate 3-D assessment of an individual sample.
These data are the first to describe the temporal and spatial relationship of intratumorally delivered IONP. We demonstrate that nanoparticles are, in fact, heterogeneously distributed on a cellular level within intra-cytoplasmic vesicles. Based on the observations made in this study, there does not appear to be a specific anatomic location of the nanoparticles within cells. It is well known that inter-particle interactions affect the magnetic properties of nanoparticles, which results in aggregated IONP displaying higher heating rates . These improved heating effects may lead to increased cytotoxicity and tumor control and are the focus of current studies.
We have demonstrated that core-shell magnetic iron oxide nanoparticles of average hydrodynamic diameter of approximately 100 nm and coated with dextran and rapidly internalized (90% within four hours) by tumors cells in vivo (figure 5). Once taken up by tumor cells, the nanoparticles are trafficked together into large collections. These results will inform future work using magnetic nanoparticles activated with alternating magnetic fields to treat tumors.
This work was supported by NIH NCI grant 1U54CA151662-01 and the Dartmouth Center of Cancer Nanotechnology Excellence (DCCNE). R. Ivkov was formerly employed by Triton BioSystems (now Aspen MediSys). The P. J. Hoopes laboratory shared a grant with Triton BioSystems (Award Number TSI-4029-08-78777) and received nanoparticles from that company to complete these studies. Triton BioSystems did not participate in any way in the production of this manuscript. A.J. Giustini gratefully acknowledges support from the Department of Education Graduate Assistantship in Areas of National Need (GAANN) Fellowship and the Thayer School of Engineering Innovation Fellowship.
The authors thankfully acknowledge Katherine S. Connolly and Christopher O. Ogomo for their assistance with TEM imaging and Charles P. Daghlian for fruitful discussions about TEM image analysis.
†now Aspen MediSys, LLC.