Modification of Aminosilanized Superparamagnetic Nanoparticles ...

B Academy of Molecular Imaging, 2009 Published Online: 7 July 2009

Mol Imaging Biol (2010) 12:25Y34 DOI: 10.1007/s11307-009-0237-9

RESEARCH ARTICLE

Modification of Aminosilanized Superparamagnetic Nanoparticles: Feasibility of Multimodal Detection Using 3T MRI, Small Animal PET, and Fluorescence Imaging Lars Stelter,1 Jens G. Pinkernelle,1 Roger Michel,1 Ruth Schwartländer,3 Nathanael Raschzok,3 Mehmet H. Morgul,3,5 Martin Koch,2 Timm Denecke,7 Juri Ruf,1 Hans Bäumler,4 Andreas Jordan,6 Bernd Hamm,1 Igor M. Sauer,3 Ulf Teichgräber1 1

Klinik für Strahlenheilkunde, CC6, Charité Campus Virchow-Klinikum, Universitätsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, Germany 2 Institut für Pathologie, CC5, Charité Campus Mitte, Universitätsmedizin Berlin, Charitéplatz 1, 10117 Berlin, Germany 3 Klinik für Allgemein-, Visceral- und Transplantationschirurgie, Experimentelle Chirurgie, CC8, Charité Campus Virchow-Klinikum, Universitätsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, Germany 4 Institut für Transfusionsmedizin, CC14, Charité Campus Mitte, Universitätsmedizin Berlin, Charitéplatz 1, 10117 Berlin, Germany 5 Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey 6 MagForce Nanotechnologies AG, Berlin, Germany 7 Klinik für Radiologie und Nuklearmedizin, Otto-von-Guericke-Universität Magdeburg, Leipziger Str. 44, 39120 Magdeburg

Abstract Purpose: The aim of our study was to modify an aminosilane-coated superparamagnetic nanoparticle for cell labeling and subsequent multimodal imaging using magnetic resonance imaging (MRI), positron emission tomography (PET), and fluorescent imaging in vivo. Procedures: We covalently bound the transfection agent HIV-1 tat, the fluorescent dye fluorescein isothiocyanate, and the positron-emitting radionuclide gallium-68 to the particle and injected them intravenously into Wistar rats, followed by animal PET and MRI at 3.0 T. As a proof of principle hepatogenic HuH7 cells were labeled with the particles and observed for cell toxicity as well as detectability by MRI and biodistribution in vivo. Results: PET imaging and MRI revealed increasing hepatic and splenic accumulation of the particles over 24 h. Adjacent in vitro studies in hepatogenic HuH7 cells showed a rapid intracellular accumulation of the particles with high labeling efficiency and without any signs of toxicity. In vivo dissemination of the labeled cells could be followed by dynamic biodistribution studies. Conclusions: We conclude that our modified superparamagnetic nanoparticles are stable under in vitro and in vivo conditions and are therefore applicable for efficient cell labeling and subsequent multimodal molecular imaging. Moreover, their multiple free amino groups suggest the possibility for further modifications and might provide interesting opportunities for various research fields. Key words: Aminosilane, Nanoparticle, Multimodal molecular imaging in vitro and in vivo, MRI, PET

Correspondence to: Lars Stelter; e-mail: [email protected]

26

L. Stelter et al.: Multimodal Modifications of Aminosilanized Nanoparticles

Introduction

I

n vivo cell trafficking has been attempted in many different experimental settings. Traditionally, cells have been followed by technically challenging and invasive chamber models using intravital microscopy or by using flow cytometry analysis of fluorescently labeled cells recovered from excised tissues [1, 2]. Molecular imaging, i.e., the visualization of single cells and cell clusters in vivo, enables tracking of cells after being applied or transplanted into an animal model. In particular, the traceability of cell migration and depiction of cellular engraftment and its affection by cellular regulatory processes by non-invasive advanced imaging technologies might facilitate new insights into multiple coherences in basic research as well as in clinical issues in medicine. Magnetic resonance imaging (MRI) with its high spatial and temporal resolution and positron emission tomography (PET) with its distinguished sensitivity provide powerful tools for molecular and cellular imaging [3, 4]. The use of fluorochromes is also an established technique to visualize cellular processes in vivo or to display localization of labeled cells using fluorescence microscopy [5, 6]. In this study, our aim was to synthesize and modify a superparamagnetic iron oxide nanoparticle for cell labeling and subsequent multimodal imaging in vivo. We favored a particle with an aminosilane coating. It exhibits multiple NH2 groups, which serve as the reactive group, and enable the covalent linkage of peptides or radioisotopes. Our project comprised basically four phases. First, we modified our superparamagnetic nanoparticles for multimodal imaging using animal PET, high-field 3-T MRI, and fluorescence microscopy. Second, we imaged these particles after intravenous (i.v.) application into a rat model and evaluated their biocompatibility, detectability, and stability as well as their distribution pattern under in vivo conditions. Third, we labeled hepatogenic HuH7 cells with the modified particles and determined their impact on cell viability and, fourth, subsequently applied the labeled cells in vivo and assayed their biodistribution over time.

Materials and Methods Synthesis of tat FITC-Modified Superparamagnetic Iron Oxide Particles Five hundred microliters of polyethyleneglycol (PEG)-coated MagForce® iron oxide nanoparticles (0.7 mol/l iron content; MagForce Nanotechnologies GmbH, Berlin, Germany), which feature an aminosilane coating and a mean size of 100 nm (core diameter 10 nm), was given to 50 mg of the transfection agent HIV1 tat linked to the fluorescent dye fluorescein isothiocyanate (FITC; Biosyntan, Gesellschaft für bioorganische Synthese mbH, Berlin), dissolved in 200 µl dimethylsulfoxide (DMSO). After that, 10 mg of N-hydroxysuccinimide (Sigma-Aldrich), dissolved in 200 µl phosphate-buffered saline (PBS), and 80 mg EDAC N-(3-dimethyaminopropyl)-N′-ethyl-carbodiimide hydrochloride (Sigma-

Aldrich), dissolved in 200 µl DMSO, were added and stirred overnight at room temperature. For purification, the reaction mixture was transferred to a PD-10 desalting column (Amersham Biosciences AB, Sweden), eluted with PBS, and collected in 500 µl fractions. Iron content of the different fractions was represented by increased darkening. These iron containing fractions were pooled and revealed an iron concentration of 5.7 mg/ml, as determined by the Dr. Bruno Lange cuvette test (No. LCK 321, measuring range of 0.2–6.0 mg/ml total iron). Briefly, 10 µl of the particles was added to 100 µl of 40% sulfuric acid and a pinch of sodium peroxodisulfate. After 30 min at 95°C, the reaction mixture was neutralized with sodium hydroxide to a final volume of 10 ml, followed by photometric estimation of the iron content. These particles were used for the following in vitro and in vivo studies.

Radiolabeling with Gallium-68 for In Vivo Application For functional imaging, the HIV-1 tat FITC iron oxide nanoparticles were labeled using the positron-emitting radionuclide gallium-68, obtained from a Ge-68/Ga-68 generator (Russia, Obninsk, 1.1 GBq). The elution and further processing has been described in detail by Zhernosekov et al. [7]. To 130 MBq eluted Ga-68, we added 500 µl of nanoparticles and 200 µl PBS and kept it at 70°C for 20 min. Ga-68 links to the amino groups of the aminosilane coating of the particles via a coordination bond. To secure the stability of the nanoparticle complex, 200 µl of the competing chelating agent diethylenetriaminepentaacetic acid (DTPA; 1 µmol) was added and incubated for 15 min at 40°C. Another purifying step by gel chromatography was performed and concordant to the particle content of the eluate a radioactive signal was detected using an ionization chamber (Capintec Radioisotope Calibrator CRC-12, Capintec Inc). A small fraction of Ga-68 bound to DTPA eluted after the nanoparticle complex (see Fig. 1). Finally, the four fractions with the highest radioactivity bound to the particles were pooled reaching a final volume of 2.0 ml and an activity of 95 MBq. A probe for quality control was withdrawn in which radiochemical purity exceeded 95%. Briefly, DTPA was added to the pooled fractions to form a water-soluble complex with all free radionuclides. After another elution with PBS on a PD-10 column, the DTPA complex eluted after the nanoparticles. Worth mentioning is that a small amount of radioactivity could not be eluted from the column and thus was not taken into consideration. Iron concentration could be defined as 1.2 mg/ml in 24 MBq/ml Ga-68 (method described above). Having a molecular weight of iron at 55.847 g/M, the specific activity could be estimated as 3.58×108 Bq/nM.

Radiolabeling with Indium-111 for In Vitro Application For in vitro use and biodistribution studies of labeled cells, 500 µl of the tat FITC nanoparticles and 200 µl PBS were added to indium-111 (In-111-chloride in diluted hydrochloric acid, Mallinckrodt Baker B.V., Netherlands) for 20 min at 70°C. After purification (see above), the three darkest fractions were pooled and contained about 90% (41 MBq) of the applied In-111 and an iron concentration of 2.9 mg/ml, resulting in a specific activity of 1.69× 108 Bq/nM.


27

Fig. 1. Elution profile of Ga-68-labeled modified nanoparticles. Fractions 1–4 showed concordant darkening, expressed by the brown box, indicating iron oxide content. In these fractions, the detected radioactive signal shows that the linkage of the radioisotope to the particles was successful. Only a small amount of radioactivity could be separated from the particles using the competing chelating agent DTPA (fractions 9–11).

Cell Survival (MTT) Assay The human hepatoma cell line HuH7, obtained from the American Type Culture Collection (ATCC®, Wesel, Germany), was incubated in DMEM medium (Biochrom AG, Berlin, Germany), supplemented with 10% fetal bovine serum, sodium pyruvate, Lglutamine, and penicillin–streptomycin, at 37°C and 5% CO2 until formation of a confluent layer. Subsequent to trypsinization, 5×104 HuH7 cells were plated in 96-well plates (BD Biosciences, Bedford, MA, USA) in triplicates. Twelve hours later, the cells were incubated with 70 µl of the modified nanoparticles in three different dilutions (A, 5.7 mg/ml; B, 2.7 mg/ml; and C, 1.5 mg/ml iron) for 2, 24, or 48 h. After removing the particle suspension, 3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromid (MTT, 5 mg/ml PBS, Roche, Mannheim, Germany) was diluted 1:10 in cell culture medium, and 200 µl was placed in each well and incubated at 37°C and 5% CO2 for 4 h. The mitochondrial reduction of the MTT salt to formazan by metabolic active cells was scanned by a multiwell spectrophotometer (ELISA reader) at 550 nm. The reference wavelength was 620 nm.

Fluorescence Microscopy HuH7 cells (5×104) were seeded into fluorescence plates (BD Biosciences) and incubated with 70 µl of modified particle suspension B (2.7 mg/ml iron) for 1 h. After a washing step, cells were fixed with 400 µl paraformaldehyde (4%) per well. Subsequently, cells were permeabilized using 400 µl methanol (80%) for 20 min, followed by washing with PBS and 1 h blocking of unspecific binding using 3% bovine serum albumin, 0.2% fish gelatin (both Sigma, Munich, Germany), and 2% fetal calf serum (Biochrom AG). After that, each plate was stained with 4′,6diamidino-2-phenylindoledihydrochloride (DAPI; Sigma, Munich, Germany) and TRITC-phalloidin (Sigma, Munich, Germany) in blocking buffer for another hour. Finally, cells were washed three times with PBS for 5 min, mounted in Aqua–Poly/Mount (Polysciences Inc., Warrington, PA, USA) and subjected to fluorescence microscopy. Fluorescence images were collected using a ×320 magnification on a dual-channel fluorescence microscope (Axiovert 40CFL, Carl Zeiss, Oberkochem, Germany). Images were taken using a cooled charged-coupled device (Photometrics, Tucson, AZ, USA) with appropriate excitation and

emission filters (Omega Optical, Brattleboro, VT, USA). The images were taken using a QICAM FAST 1394 Camera (QIMAGING, Surray, BC, Canada) and processed using Confocal Assistant.

Magnetic Resonance Imaging In Vitro HuH7 cells (2.5×105) were seeded into six-well plates (Sarstedt, Nümbrecht, Germany) and, after reaching confluency, incubated with the modified particles for 1 h in the following two groups: (A) control, just 1 ml medium per well, and (B) 1 ml particle suspension B per well. For sample preparation, 1 g agarose in 100 ml distilled water was heated up to 80°C and 200 µl was given to 8×105 cells in a small test tube. To avoid future artefacts in MR imaging, air bubbles had to be excluded and a homogeneously distributed suspension had to be secured before cooling down. In that condition, agarose provides a 3-D matrix which provides good dispersion and immobilization of the labeled cells. The cell-agarose test tubes underwent high-resolution imaging on a clinical MR scanner at a field strength of 3.0 T (Signa 3T94, GE Healthcare, Milwaukee, WI, USA). The maximum gradient amplitude was 40 mT/m. A circularly polarized surface coil with a diameter of 2 cm was used (Rapid Biomedical, Würzburg, Germany), applying a T2*-weighted, 2-D gradient-echo pulse sequence [8]. In detail, imaging was performed at a repetition time (TR) of 200 ms, an echo time (TE) of 25 ms, and a flip angle (α) of 20°. Nominal resolutions of 78 × 78 × 800 µm3 and 39 × 39 × 800 µm3 were achieved from a field of view (FOV) of 20 mm, a matrix size of 256×256 and 512×512, respectively, and a slice thickness of 0.8 mm. The number of excitations acquired (NEX) was kept at 12, resulting in a total imaging time of about 5 min for single slice acquisitions.

Animals Female Wistar rats (200–250 g; Charles River WIGA GmbH, Sulzfeld, Germany) were anesthetized with a combination of Ketanest® (ketamine hydrochloride, 100 mg/ml, Parke-Davis GmbH, Karlsruhe, Germany) and Rompun® (xylazine hydrochloride, 20 mg/ml, Bayer AG, Leverkusen, Germany) at a dose of 200 and 10 µg per gram body weight by intraperitoneal injection.

28


Animal care followed institutional guidelines, and all experiments were approved by local animal research authorities.

(Wallac 1480 Wizard™ 3″). Results were calculated as percent injected dose per gram (%ID/g) of tissue.

Animal PET Imaging

Results

For PET imaging, the animals (n=3) were anesthetized followed by an intravenous injection of 10 MBq (about 0.48 mg iron in 400 µl total volume) of the modified and Ga-68-labeled nanoparticles through a lateral tail vein. PET imaging was performed using a dedicated small bore animal PET scanner (Philips Mosaic™, Philips Medizin Systeme GmbH, Hamburg, Germany). This device uses a 2×2×103-mm gadolinium-oxyorthosilicate optically continued pixelated Anger-logic detector with conventional 19 mm diameter photomultiplier tubes leading to high-resolution (approx. 2 mm), high-sensitivity and high-counting rate. The detector ring has a 20.8 cm diameter, the gantry a 16-cm aperture and 12.8 cm transverse FOV and operates exclusively in 3-D volume imaging mode. Animals were imaged starting with six series of 5 min and a following emission scan of 30 min. Images were reconstructed in cubic voxels of dimension 0.5 mm3 using 3-D-RAMLA.

Animal MR Imaging Each animal underwent MR imaging 1 and 24 h post-injection using a T2-weighted fast recovered fast spin echo (T2w-frFSE) sequence at a clinical 3 T MRI scanner (GE Signa 3T94™, Milwaukee, WI, USA). The imaging parameters were as follows: TR/TE=2,000/85 ms, echo train length 16, NEX 4–8, slice thickness of 0.5–1.0 mm, and a 256×256 or 512×512 matrix using a FOV of 4–10 cm, depending on slice orientation. A dedicated small animal two-channel quadrature volume resonator (Biomedizinische Geräte GmbH Würzburg, Germany) with an inner diameter of 3.5 cm was used to yield high temporal and spatial resolution. Prior to imaging, a manual adaption of the coil frequency and the impedance of the system to 50 Ω were performed using a RF sweeper (405NV, Morris Instruments Inc., Ottawa, ON, Canada). Images were obtained in coronal and axial slicing within approximately 15 min per animal.

Biodistribution Studies After sacrificing the animals, intracellular iron content of the liver and spleen was analyzed by Prussian blue staining. In brief, 7 µm thick cryosections were fixed in 4% paraformaldehyde in PBS, pH 7.0, for 20 min at room temperature and stained with 5% potassium ferrocyanide and 5% hydrochloric acid. Sections were examined with an optical microscope (Axiovert 25, Carl Zeiss, Jena, Germany) at ×100/×200/×400 magnification. To evaluate tissue distribution of labeled HuH7 cells in vivo, 2× 106 cells were seeded into 25 cm2 flasks (Sarstedt) and, after reaching confluency, were incubated with 500 µl of the In-111labeled particles (2.9 mg/ml iron content) as described above. After trypsinization, approximately 1×106 cells (about 3 MBq) were injected i.v. and the animals were sacrificed after 2, 24, or 48 h (n= 3, for each group). During the post-injection period, no adverse reactions or signs of discomfort and sickness were observed. After anesthesia and cervical dislocation, blood was collected, and the organs were dissected, washed free of blood, and weighed in an analytic balance. Associated radioactivity was counted in an appropriate energy window (±20%) in an automatic γ-counter

Radiolabeling of Modified Nanoparticles The multiple amino groups featured by the particles’ aminosilane coating enabled us to covalently link the HIV1 tat FITC peptide as well as the positron-emitting radionuclide gallium-68 or the γ-emitter indium-111. The purification step of the Ga-68-labeled particles is exemplarily shown in Fig. 1. The y-axis of the graph represents the activity of each 500 µl fraction measured in a γ-counter and the x-axis shows the eluted fractions, numbered 1–15. In this demonstrated elution profile, a constant darkening of fractions 1–4 (brown area) represented iron content and also showed a concordant radioactive signal, indicating that the linkage was successful. To secure stability of the synthesized complex, the competing chelator DTPA was added. However, only a small amount of radioactivity (fractions 9–11, Fig. 1) was separated from the nanoparticles by DTPA, which eluted later than the complex due to its smaller size. This indicates that radioactive labeling of the particles was not only successful but also stable. The final particle size showed to be basically unchanged with about 100 nm as detected by dynamic light scattering (ZetaSizer 3000HS, Malvern Instruments Ltd, UK).

Intracellular Uptake and Cell Viability Efficient labeling of HuH7 cells with our modified nanoparticles could be obtained within 1 h. Particle load did basically not alter the cells viability. After incubating different concentrations of particles over a period of up to 48 h, HuH7 cells showed a similar metabolic activity compared to the untreated control, as demonstrated by the MTT assay (Fig. 2). Additionally, and as displayed by fluorescence microscopy in Fig. 3, functional integrity of the cells could be maintained, showing a dividing cell in the center of the image. The cells actin filaments are nicely aligned in parallel fibers, do not show any abruption, and are located in all adhesion areas. Staining with DAPI also showed no fragmentation of the nuclei. The majority of cells presented a considerable green signal emitted from the particles fluorochrome, demonstrating their intracellular localization.

Magnetic Resonance Imaging In Vitro To see if intracellular particle load could also be detected by MRI, we performed a high-resolution MRI of labeled HuH7 samples dispersed in an agarose matrix. Compared with the control preparations of agarose medium alone (data not shown) or with non-labeled cells (Fig. 4a), the labeled cells could be displayed as distinct signal


29

Fig. 2. MTT proliferation assay of HuH7 cells incubated with a serial dilution of the modified nanoparticles over a period of 48 h. Besides the early time point (2 h) and the group receiving high iron concentrations (5.7 mg/ml), no significant reduction in cell viability seems to be caused by particle load.

extinctions of 200–400 µm in diameter. These signals are due to the susceptibility artefact given by iron oxide (Fig. 4b, black arrowhead) and clearly contrast the high signal of the aqueous agarose background. An increase in labeled cells correlated with an ascending number of extinctions observed in MRI. Subsequent analysis of the samples with optical microscopy showed the cell population to be monodisperse with only a few cellular aggregates (images not shown). Therefore, the majority of extinctions represent individual cells loaded with different amounts of iron. However, a few aggregates can be distinguished as larger sized signal extinctions (white arrowhead).

PET and MRI In Vivo After intravenous injection of about 10 MBq of our particle suspension into Wistar rats (n=3), PET imaging was performed over the first hour post-injection. A distinct radionuclide distribution in the upper abdomen could already be detected after 5 min and increased during the next 50 min. Apart from the liver and a faint accumulation in the spleen, no other region showed a radioactive signal (Fig. 5d). Subsequent to PET, MR imaging was performed after different periods of time. As demonstrated in transaxial slices of the upper abdomen in Fig. 5b, a slight hypointensity of the liver

Fig. 3. Fluorescent staining of labeled HuH7 hepatoma cells. Staining of actin filaments was reached using phalloidin (red signal), DAPI for staining of the nuclei (blue signal), and the green fluorescent signal evoked by FITC-labeled nanoparticles. Magnification, ×320.

30


Fig. 4. High-resolution MR images of particle-loaded HuH7 cell samples dispersed in an agarose matrix. Compared with nonincubated cells (a), distinct punctuate signal extinctions caused by the iron oxides’ susceptibility artefacts could be observed in labeled single cells (b, black arrowhead), respectively cell clusters (b, white arrowhead).

Fig. 5. Coronal PET after i.v. application of 10 MBq-modified nanoparticles (d). The orange arrows indicate liver accumulation of the particles. Notably, there was no other region (apart from the injection side at the tail, red arrowhead, and faintly in the region of the spleen, green arrowhead), which showed a signal in PET or MRI. The lack of signal in the left upper abdomen in the PET scan (yellow asterisk) is caused by air and/or food in the stomach. T2-frFSE-weighted MR imaging of the liver in axial slicing before (a), 1 h after (b), and 24 h after (c) i.v. application of the modified nanoparticles. A commencing alteration of the liver signal can be observed, followed by increasing hypointensity of the liver parenchyma due to accumulation of the iron oxide particles in the RES. Axial PET image of the liver 50 min post-injection (f) showed a clear radioactive signal throughout the whole organ. Successful bimodal imaging of the particles in vivo underlined by PET (f) and MR image (e) fusion (g).


after 1 h and a significant signal drop after 24 h (Fig. 5c) could be detected when compared to the control (Fig. 5a), indicating a commencing iron oxide content. Anatomic correlation of the PET signal and the findings in MRI could be verified by image fusion (Fig. 5e–g). Besides the localizations named above, no activity retention could be seen in the kidneys or the urinary bladder, underlining that the synthesis and modification processes were successfully performed and proved to be stable under in vivo conditions.

31

Histological Work-up Histological preparation of the organs with indication of particle accumulation was performed as a following step. Prussian blue staining revealed iron oxide content in the reticuloendothelial system (RES) located in the liver sinusoids and the spleen (black arrows, Fig. 6a, c). Sufficient and stable labeling of the particles with the fluorochrome could be demonstrated by an analogous green signal in fluorescence microscopy (Fig. 6b).

Fig. 6. Histology of the liver and spleen after in vivo imaging. Prussian blue staining reveals iron content (black arrows) in the liver sinusoids (a, magnification ×400) and the RES of the spleen (c, magnification ×400). b A comparable liver section using fluorescence microscopy (fluorescent particle marked by a white arrow).

32


Biodistribution of Labeled Cells In Vivo As a proof of principle, HuH7 cells were labeled with the modified nanoparticles, carrying the radioisotope In-111, and about 1×106 cells were injected i.v. into each rat (n=9) through the retro-orbital sinus. The animals were randomly assigned to three groups, anesthetized, and sacrificed 2, 24, and 48 h post-injection. The tissue distributions of the labeled cells were determined by γ-counter measurements and are presented in Table 1. Interestingly, the pattern of cellular dissemination is different than that of the nanoparticles alone. In the first 2 h, the labeled cells are carried through the bloodstream into the lungs. Within 48 h, the cells get washed out of the lung and blood and distribute evenly throughout the body. A certain accentuation of cellular accumulation can be seen in the kidneys, presumably in the glomerular system, and the spleen.

Discussion The most promising and so far the best-studied technique to non-invasively visualize cells under in vivo conditions is based on superparamagnetic Fe2O3 particles and MRI [9, 10]. MRI offers unique spatial resolution with outstanding contrast features in parenchymal organs. There are different contrast techniques depending on T1 relaxation or T2 relaxation times, which are adjustable by magnetic contrast material. Nanoscaled iron oxide particles strongly promote dephasing (T2 relaxation) of proton spins and therefore cause signal distinctions in MRI. Depending on chemical features of their polymeric coating, some iron oxide nanoparticles are feasible for chemical modifications. In this way, their cellular uptake rate may be increased as well as their utilization for other imaging modalities based on fluorochromes or radioisotopes. In our study, we used MagForce® nanoparticles as a superparamagnetic vehicle, consisting of a 10 nm Fe2O3 nucleus embedded by aminosilane coating [11]. These particles exhibit multiple NH2 groups, bound to a silicon atom, serving as the reactive group and enabling the covalent linkage of peptides or fluorescent dyes [9]. In our case, we used the polycationic peptide HIV-1 tat and the

fluorochrome FITC as well as the positron-emitting radionuclide gallium-68 and the γ-emitter indium-111 without adding a chelating agent, such as DTPA or DOTA. An additional PEG coating was used to block adjacent amino functions and prevent particle aggregation. To get an idea of how these modified nanoparticles will distribute in an animal and if they will be stable under in vivo conditions, we first applied them intravenously in a healthy rat model. Subsequently, we performed small animal PET followed by MR imaging and we were able to noninvasively detect our particles basically in the liver and spleen with both modalities. It is well known that iron oxide accumulates during first pass in the reticuloendothelial system of the liver and the spleen [12, 13]. Enabled by the added FITC, fluorescence microscopy revealed the histological affirmation. However, no other significant nanoparticle accumulation could be observed, especially in the PET studies. In particular, no activity retention could be seen in the kidneys or the urinary bladder, through which most of the radioligands are predominantly excreted. These concordant findings led to the conclusion that our modifications were successful and proved to be stable under in vivo conditions. One possible application for our modified particles could be ex vivo labeling of cells destined for in vivo application. Although aminosilane coating itself seems to mediate intracellular uptake in certain cell lines [8, 14], we also colinked the peptide HIV-1 tat to the particle. Tat peptides mediate cell surface adherence and internalization by an adsorptive endocytosis process through cellular and nuclear membranes [15]. When covalently coupled to iron oxide particles, it allows highly efficient cellular transfection, prolonged intracellular retention without significantly affecting cellular integrity and function [12, 16–18]. Alternatively, the FDA-approved drug protamine might be used to enhance cellular particle load and might be favorable in possible prospective applications of nanoparticles in translational studies, due to its low costs and known biological effects in a clinical setting [19]. As a proof of principle for in vitro use of our particles, we performed incubation studies on human hepatogenic HuH7

Table 1. Biodistribution of HuH7 cells labeled with In-111-modified nanoparticles, estimated by γ-counter 2, 24, and 48 h post-injection %ID/g (mean ± SD) 2 h (n=3) Blood Heart Lung Liver Spleen Kidneys Bladder Intestine Stomach

4.67±0.52 2.59±0.04 14.98±1.22 2.18±0.05 1.46±0.02 4.67±0.08 3.24±0.07 0.37±0.02 0.026±0.03

Values are expressed in percent injected dose per gram tissue (%ID/g) (mean ± SD)

24 h (n=3) 1.82±0.28 1.3±0.28 3.24±0.85 2.74±0.29 4.24±0.21 6.3±0.35 1.61±0.36 0.67±0.16 0.35±0.11

48 h (n=3) 0.98±0.24 1.16±0.4 1.16±0.13 2.33±0.2 5.06±0.25 7.91±0.34 0.53±0.24 0.66±0.12 0.43±0.15


cells. Using fluorescence microscopy and photometric measuring of cell lysates, we were able to observe a clear intracellular localization of the particles within 1 h. No signs of cellular impairment or loss of integrity could be seen. In a proliferation assay performed after incubation of a serial dilution of our nanoparticle suspension over a period of 48 h, only a slight decline of cell viability occurred after 2 h within the highest iron concentration group (5.7 mg/ml). Looking at 24 and 48 h of cell labeling, this effect seemed to be inverted towards superior proliferation capability compared to the controls. However, this has been observed by others before [20] and is presumably due to an alteration of the assays’ extinction caused by the intracellular iron load. The influence of intracellular particle load should be assayed individually for each cell line looking at cell specific functions and estimation of toxicity. Applying a T2*-weighted sequence in MRI, we were able to image the labeled HuH7 cells as single cells and cell clusters in an agarose matrix. However, after injecting the labeled cells in rats, MR imaging failed to detect cellular localization in vivo (images not shown). Looking at the biodistribution pattern, obtained by the radioactive signal emitted from the particles and listed in Table 1, we saw an initial predominant localization of the labeled cells in the lungs. Further on, the cells seemed to more or less evenly distribute throughout the animals’ body and the amount of iron oxide per cell might not have been enough to get a considerable change in MRI signal. Here, the application of the more sensitive PET might help to detect cells in vivo, when they were given systemically and not “unilocular” to a specific organ. Also, labeling cells with a more specific “target” such as macrophages or other cells activated by, e.g., inflammation [17, 21] or applying ex vivo labeled cells to a defined region, such as the heart [22], might be a suitable application using our particles. In any case, the efficiency of labeling and the impact of the particles on each individual cell line will have to be determined prior to its use in vivo as well as the specific iron content per single cell for in vivo MR detection has to be defined. Radiolabeling of nanoparticles has been described in some rare cases using, e.g., In-111 and Cu-64 [23–25]. Using Ga-68 and In-111, we could prove that radiolabeling of our particles could easily be done without the need of a chelating agent. Ga-68 has the advantage of being relatively easy to synthesize using a Ge-68/Ga-68 generator. However, a major disadvantage is its relatively short half-life (approximately 1.13 h), limiting its use for in vivo cell trafficking by PET. In-111 with its longer half-life (2.8 days) would be more suitable to answer the question of the “fate” of cells after in vivo application, but is limited to the use of the less sensitive single photon emission computed tomography. As the aminosilane coating of our particles offer multiple free amino groups, covalently linking positron-emitting isotopes with a longer half-life, e.g., I-124 (4.2 days), seems to be possible and will enable to use the more sensitive PET. At the time of this study, I-124 was not accessible in our lab,

33

but labeling our nanoparticles with a longer lasting positronemitting isotope will be addressed in further studies. We think, due to its coating-specific characteristics, our particles provide the advantage of easy and straightforward modifications and are therefore a powerful tool for various fields in translational science.

Conclusion In conclusion, aminosilane coating seems to be favorable for multimodal linking of magnetic nanoparticles, as it provides multiple positively charged amino reactive groups. Radioisotopes or even complex substances such as fluorescent dyes or peptides can be stably linked and enable in vivo and ex vivo imaging using small animal PET, clinical 3 T MRI, and fluorescence microscopy. These imaging approaches have been successfully combined in one particle and provide an attractive tool for cell tracking in vivo. Another conceivable utilization would be its dedication in the field of cellular transplantation. Extensive laboratory studies in animal models on transplanting, e.g., syngenic or allogenic hepatocytes have been performed already [6, 24–28]. However, it is not yet fully understood how transplanted cells interact within an impaired organism and to what extent they assume functionality [29]. Our modified nanoparticles could be suitable to overcome these limitations and might therefore help to answer critical questions in a preclinical setting. References 1. Menger M.; Lehr H. Scope and perspectives of intravital microscopy bridge over from in vitro to in vivo. Immunology Today 14: 519-522, 1993. 2. Hendrikx P.; Martens A.; Hagenbeek A.; Heij J.; Visser J. Homing of fluorescently labeled murine hematopoietic stem cells. Exp. Hematol. 24: 129-140, 1996. 3. Kikkawa H, Tsukada H, Oku N. Usefulness of Positron Emission Tomographic Visualization for Examination of In vivo Susceptibility to Metastasis. Cancer. 89: 1626-33, 2000 4. Perez JM, Josephson L, Weissleder R. Use of Magnetic Nanoparticles as Nanosensors to Probe for Molecular Interactions. Chem Bio Chem. 5: 261-264, 2004. 5. Misgled T, Nikic I, Kerschensteiner M. In vivo imaging of single axons in the mouse spinal cord. Nat Protoc. 2: 263-268, 2007. 6. Kaufman CL, Williams M, Ryle LM, Smith TL, Tanner M, Ho C. Superparamagnetic iron oxide particles transactivator protein-fluorescein isothiocyanate particle labeling for in vivo magnetic resonance imaging detection of cell migration: uptake and durability. Transplantation. 76: 1043-1046, 2003. 7. Zhernosekov KP, Filosofov DV, Baum RP, et al. Processing of generator-produced 68Ga for medical application. J Nucl Med. 48 (10): 1741-8, 2007. 8. Koch A, Reynolds F, Kircher M, Merkle H, Weissleder R, Josephson L. Uptake and Metabolism of a Dual Fluorochrome Tat-nanoparticle in HeLa Cells. Bioconjugate Chem. 14: 1115-1121, 2003. 9. Josephson L, Tung C, Moore A, Weissleder R. High-efficiency intracellular magnetic labeling with novel superparamagnetic-Tat peptide conjugates. Bioconjugate Chem. 10: 186-91, 1999. 10. Jordan A, Scholz R, Maier-Hauff K, et al. The effect of thermotherapy using magnetic nanoparticles on rat malignant glioma. J NeuroOncology. 78: 7-14, 2006. 11. Wunderbaldinger P, Josephson L, Weissleder R. Tat Peptide Directs Enhanced Clearance and Hepatic Permeability of Magnetic Nanoparticles. Bioconjugate Chem. 13: 264-268, 2002.

34


12. Shenoy D, Little S, Langer R, Amiji M. Poly(ethylene oxide)-modified poly(beta-amino ester) nanoparticles as a pH-sensitive system for tumor-targeted delivery of hydrophobic drugs: part 2. In vivo distribution and tumor localization studies. Pharm Res. 22: 2107-14, 2005. 13. Pinkernelle J, Teichgräber U, Neumann F, et al. Imaging of Single Human Carcinoma Cells in vitro Using a Clinical Whole-Body Magnetic Resonance Scanner at 3.0 T. Magnetic Resonance in Medicine. 53: 1187-1192, 2005. 14. Ma YJ, Gu HC, Study on the endocytosis and the internalization mechanism of aminosilane-coated Fe3O4 nanoparticles in vitro. J Mater Sci Mater Med. 18: 2145-2149, 2007. 15. Lundberg M, Wikström S, Johansson M. Cell Surface Adherence and Endocytosis of Protein Transduction Domains. Molecular Therapy. 8: 143-150, 2003. 16. Fawell S, Seery J, Daikh Y, et al. Tat-mediated delivery of heterologous proteins into cells. Proc. Natl. Acad. Sci. 91: 664-668, 1994. 17. Pittet M, Swirski F, Reynolds F, Josephson L, Weissleder R. Labeling of immune cells for in vivo imaging using magnetofluorescent nanoparticles. Nature Protocols. 1: 73-79, 2006. 18. Becker M, Bailey L, Wooley K. Peptide-derivatized shell-cross-linked nanoparticles. 2. Biocompatibility evaluation. Bioconjugate Chem. 15: 710-717, 2004. 19. Reynolds F, Weissleder R, Josephson L. Protamine as an efficient membrane-translocating peptide. Bioconjugate Chem. 16: 1240-5, 2005. 20. Gilad AA, Walczak P, McMahon MT, et al. MR Tracking of Transplanted Cells With "Positive Contrast" Using Manganese Oxide Nanoparticles. Magnetic Resonance in Medicine. 60: 1-7, 2008.

21. Moore A, Grimm J, Han B, Santamaria P. Tracking the recruitment of diabetogenic CD8+ T-cells to the pancreas in real time. Diabetes. 53: 1459-1466, 2004. 22. Rossin R, Pan D, Qi K, et al. 64Cu-Labeled Folate-Conjugated Shell Cross-Linked Nanoparticles for Tumor Imaging and Radiotherapy: Synthesis, Radiolabeling, and Biologic Evaluation. J Nucl Med. 46: 1210-1218, 2005. 23. DeNardo S.; DeNardo G.; Natarajan A. et al (2007) Thermal dosimetry predictive of efficacy of 111In-ChL6 nanoparticle AMF-induced thermoablative therapy for human breast cancer in mice. J Nucl Med. 48: 437-444 24. Nahrendorf M, Zhang H, Hembrador S, et al. Nanoparticle PET-CT Imaging of Macrophages in Inflammatory Atherosclerosis. Circulation. 117: 379-387, 2008. 25. Makowka L, Falk R, Rotstein L, et al. Cellular transplantation in the treatment of experimental hepatic failure. Science. 210: 901-903, 1980. 26. Horslen S, Fox I, Hepatocyte transplantation. Transplantation. 77: 14811486, 2004. 27. Matas A, Sutherland D, Steffes M, et al. Hepatocellular transplantation for metabolic deficiencies: decrease of plasma bilirubin in Gunn rats. Science. 192: 892-894, 1976. 28. Amsalem Y.; Mardor Y.; Feinberger M.S. et al. Iron-oxide labeling and outcome of transplanted mesenchymal stem cells in the infarcted myocardium. Circulation. 116(11 Suppl): I38-45, 2007. 29. Lee KD, Kuo TKC, Whang-Peng J, et al. In vitro Hepatic Differentiation of Human Mesenchymal Stem Cells. Hepatology. 40: 1275-1284, 2004.