Fluorescent magnetic nanoparticles for magnetically enhanced cancer imaging and targeting in living subjects.

METHODS/EXPERIMENTAL FMN preparation SPIO were synthesized by oxidative decomposition of iron pentacarbonyl in trimethylamine oxide, oleic acid and hexadecane. 24 The average particle size is 8 nm, as measured by transmission electron microscopy (TEM) ( Supplementary Figure 4 ). Afterward, silanization coating chemistry was developed to render SPIO water soluble and biocompatible. Briefly, mercapto/amino propyl trimethoxyl silane was added to coat the surface of SPIO by forming a crosslinked shell under basic conditions using tetramethylammonium hydroxide as the base in a methanol solution. By adjusting the quantity of silane molecules added, the shell thickness can be controlled. In this work, the size of silanized SPIO before fluorophores incorporation is 87 nm as measured by Dynamic Light Scattering technique, which provide a silanization coating thickness of roughly 40 nm. Both mercapto and amino groups are incorporated onto the nanoparticle surface. Near infrared fluorophores (Cy55) with NHS ( N -hydroxysulfosuccinimide) modification (GE Healthcare) are covalently bonded to the amino groups. A small Cyclo[-Arg - Gly - Asp - D - Tyr – Lys] (RGD) that can specifically target tumor angiogenesis marker α v β 3, or its inactive control peptide Cyclo[- Arg - D - Ala - Asp - Tyr – Lys] (RAD), were bound to mercapto groups of the silanized SPIO surface using a crosslinker Sulfo-SMCC (Sulfosuccinimidyl 4-[ N -maleimidomethyl]cyclohexane-1-carboxylate, Thermo Fisher Scientific).

FMN characterization Transmission electron microscopy

(TEM) imaging and characterization was performed with using a Jeol TEM1230 at 80 KV. Dynamic Light Scattering was performed on a ZetaPlus Analyzer (Brookhaven Instruments Corporation). Animal experiments were conducted in accordance with the ethical guidelines of the National Institutes of Health and with the approval of the Institutional Animal Care and Use Committee of Stanford University. Intravital imaging was performed using an IV−100 intravital microscope (Olympus, Center Valley, PA). The experimental preparation included: (1) surgically implant a dorsal skinfold chamber (APJ Trading Co., Inc., Ventura, CA) onto the back of the SCID mouse (Charles River, Wilmington, MA) which is a retired male breeder weighing more than 28g, and wait for 2–3 days; (2) 1 million EGFP-transfected U87MG human glioblastoma cells were inoculated in low volume of 20 μl directly on top of the skinfold in the DSC. Tumor was allowed to grow for 10 days before imaging. On the day of intravital imaging, the mouse was anesthetized with isoflurane and positioned with the DSC fixed beneath the 10x objective of the intravital microscope. Commercial vascular dye Angiosense 750 (VisEn Medical, Woburn, MA) and FMN (with RGD or RAD) were subsequently injected into the mouse tail vain. Laser sources at 488 nm, 633 nm, and 748 nm were used for excitation. Three corresponding output channels, illustrated in this paper with green for the EGFP transfected tumor, red for HFMNs-Cy5.5, and blue for Angiosense 750 were simultaneously scanned for image acquisition. 320 and 512 pixel resolutions were selected in most imaging situations. The injection amount for Angiosense is 100 μl of original solution, which corresponds to 1.3 nanomoles. The FMN concentration was kept at 0.1 μM. A 100 μl solution of FMN (10 picomole) was injected for most experiments except for the dosage experiments, where 200 μl (20 picomole) of FMN were injected.

Intravital imaging permits direct observation of the behavior

FMN in living subjects. Magnetic targeting The observation of magnetic nanoparticle accumulation in vivo was accomplished using an intravital microscope and a tumor xenograft mouse model in a dorsal skinfold chamber (DSC). For a general experiment, EGF-transfected U87MG human glioblastoma cells (~ 1 × 10 6 ) were inoculated and grown on the back of one mouse, and inside the DSC, for 10 days. Then commercial vascular dye Angiosense ® 750 and FMN (with RGD or RAD) were intravenously injected, both of which outline the tumor neovasculature. The tumor area and tumor neovasculature were examined using intravital microscopy after the injection of Angiosense and FMN. Because tumor vascular growth is heterogeneous, including variations of the depth and separation of vessels, an area with the best vessel structure in the entire observation window is selected. Then a small piece (edge length ~ 3 – 8 mm) of electroformed Ni micromesh (wire width 12 μm and thickness 5 μm) was placed on top of the area. The magnetic force began when a permanent NdFeB N52 rectangular bar magnet (7.5 × 7.5 × 22 mm 3 ) was placed beneath the DSC window on the back side of the skinfold. The permanent magnet has a field strength of 2 kOe and a maximum field gradient of 10 T/m near the surface.

Statistical analysis

The statistics for magnetic targeting ( Figure 3D ) were obtained from the average and standard deviation of fluorescence intensity ratios corresponding to the three pairs of colored arrows. The statistical analysis for tumor signal change with time of FMN-RGD, FMN-RAD, and magnetic targeting ( Figure 4D ) was based on the intensity of eight ROIs from two representative images of each imaging time point. The ROIs were selected to be within regions with the brightest tumor (EGFP) signals. The statistical analysis of the half-lives of tumor signal decay ( Figure 4E ) was generated by fitting each of the six FMN-RGD injection curves ( Figure 4D and Supplementary Figure 3 ), 3 for injection under magnetic targeting and 3 for injection without magnetic targeting, to a first order exponential decay function. The statistically different tumor signal decay trends were confirmed by a student’s t-test analysis of FMN-RGD injections with and without magnetic targeting. (n=3, p